Upload folder using huggingface_hub
Browse files- chat_template.jinja +16 -0
- config.json +1657 -0
- factorized_modules.json +1570 -0
- model.safetensors +3 -0
- modeling_pmre_olmo.py +241 -0
- tokenizer.json +0 -0
- tokenizer_config.json +12 -0
chat_template.jinja
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{% set has_system = messages|selectattr('role', 'equalto', 'system')|list|length > 0 %}{% if not has_system %}{{ '<|im_start|>system
|
| 2 |
+
You are OLMo, a helpful function-calling AI assistant built by Ai2. Your date cutoff is November 2024, and your model weights are available at https://huggingface.co/allenai. You do not currently have access to any functions. <functions></functions><|im_end|>
|
| 3 |
+
' }}{% endif %}{% for message in messages %}{% if message['role'] == 'system' %}{{ '<|im_start|>system
|
| 4 |
+
' + message['content'] }}{% if message.get('functions', none) is not none %}{{ ' <functions>' + message['functions'] + '</functions><|im_end|>
|
| 5 |
+
' }}{% else %}{{ ' You do not currently have access to any functions. <functions></functions><|im_end|>
|
| 6 |
+
' }}{% endif %}{% elif message['role'] == 'user' %}{% if message.get('functions', none) is not none %}{{ '<|im_start|>user
|
| 7 |
+
' + message['content'] + '
|
| 8 |
+
' + '<functions>' + message['functions'] + '</functions><|im_end|>
|
| 9 |
+
' }}{% else %}{{ '<|im_start|>user
|
| 10 |
+
' + message['content'] + '<|im_end|>
|
| 11 |
+
' }}{% endif %}{% elif message['role'] == 'assistant' %}{{ '<|im_start|>assistant
|
| 12 |
+
' }}{% if message.get('content', none) is not none %}{{ message['content'] }}{% endif %}{% if message.get('function_calls', none) is not none %}{{ '<function_calls>' + message['function_calls'] + '</function_calls>' }}{% endif %}{% if not loop.last %}{{ '<|im_end|>' + '
|
| 13 |
+
' }}{% else %}{{ eos_token }}{% endif %}{% elif message['role'] == 'environment' %}{{ '<|im_start|>environment
|
| 14 |
+
' + message['content'] + '<|im_end|>
|
| 15 |
+
' }}{% endif %}{% if loop.last and add_generation_prompt %}{{ '<|im_start|>assistant
|
| 16 |
+
<think>' }}{% endif %}{% endfor %}
|
config.json
ADDED
|
@@ -0,0 +1,1657 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"transformers_version": "5.9.0",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Olmo3ForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"output_hidden_states": false,
|
| 7 |
+
"return_dict": true,
|
| 8 |
+
"dtype": "bfloat16",
|
| 9 |
+
"chunk_size_feed_forward": 0,
|
| 10 |
+
"is_encoder_decoder": false,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "LABEL_0",
|
| 13 |
+
"1": "LABEL_1"
|
| 14 |
+
},
|
| 15 |
+
"label2id": {
|
| 16 |
+
"LABEL_0": 0,
|
| 17 |
+
"LABEL_1": 1
|
| 18 |
+
},
|
| 19 |
+
"problem_type": null,
|
| 20 |
+
"vocab_size": 100278,
|
| 21 |
+
"hidden_size": 4096,
|
| 22 |
+
"intermediate_size": 11008,
|
| 23 |
+
"num_hidden_layers": 32,
|
| 24 |
+
"num_attention_heads": 32,
|
| 25 |
+
"num_key_value_heads": 32,
|
| 26 |
+
"hidden_act": "silu",
|
| 27 |
+
"max_position_embeddings": 65536,
|
| 28 |
+
"initializer_range": 0.02,
|
| 29 |
+
"use_cache": false,
|
| 30 |
+
"pad_token_id": 100277,
|
| 31 |
+
"bos_token_id": null,
|
| 32 |
+
"eos_token_id": 100257,
|
| 33 |
+
"tie_word_embeddings": false,
|
| 34 |
+
"rope_parameters": {
|
| 35 |
+
"attention_factor": 1.2079441541679836,
|
| 36 |
+
"beta_fast": 32.0,
|
| 37 |
+
"beta_slow": 1.0,
|
| 38 |
+
"factor": 8.0,
|
| 39 |
+
"original_max_position_embeddings": 8192,
|
| 40 |
+
"rope_type": "yarn",
|
| 41 |
+
"rope_theta": 500000
|
| 42 |
+
},
|
| 43 |
+
"attention_bias": false,
|
| 44 |
+
"attention_dropout": 0.0,
|
| 45 |
+
"rms_norm_eps": 1e-06,
|
| 46 |
+
"sliding_window": 4096,
|
| 47 |
+
"layer_types": [
|
| 48 |
+
"sliding_attention",
|
| 49 |
+
"sliding_attention",
|
| 50 |
+
"sliding_attention",
|
| 51 |
+
"full_attention",
|
| 52 |
+
"sliding_attention",
|
| 53 |
+
"sliding_attention",
|
| 54 |
+
"sliding_attention",
|
| 55 |
+
"full_attention",
|
| 56 |
+
"sliding_attention",
|
| 57 |
+
"sliding_attention",
|
| 58 |
+
"sliding_attention",
|
| 59 |
+
"full_attention",
|
| 60 |
+
"sliding_attention",
|
| 61 |
+
"sliding_attention",
|
| 62 |
+
"sliding_attention",
|
| 63 |
+
"full_attention",
|
| 64 |
+
"sliding_attention",
|
| 65 |
+
"sliding_attention",
|
| 66 |
+
"sliding_attention",
|
| 67 |
+
"full_attention",
|
| 68 |
+
"sliding_attention",
|
| 69 |
+
"sliding_attention",
|
| 70 |
+
"sliding_attention",
|
| 71 |
+
"full_attention",
|
| 72 |
+
"sliding_attention",
|
| 73 |
+
"sliding_attention",
|
| 74 |
+
"sliding_attention",
|
| 75 |
+
"full_attention",
|
| 76 |
+
"sliding_attention",
|
| 77 |
+
"sliding_attention",
|
| 78 |
+
"sliding_attention",
|
| 79 |
+
"full_attention"
|
| 80 |
+
],
|
| 81 |
+
"_name_or_path": "allenai/OLMo-3-7B-Think",
|
| 82 |
+
"model_type": "olmo3",
|
| 83 |
+
"output_attentions": false,
|
| 84 |
+
"auto_map": {
|
| 85 |
+
"AutoModelForCausalLM": "modeling_pmre_olmo.PmreOlmoForCausalLM"
|
| 86 |
+
},
|
| 87 |
+
"pmre_metadata": [
|
| 88 |
+
{
|
| 89 |
+
"module_path": "model.layers.0.self_attn.q_proj",
|
| 90 |
+
"in_features": 4096,
|
| 91 |
+
"out_features": 4096,
|
| 92 |
+
"rank": 512,
|
| 93 |
+
"sparse_nnz": 0
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"module_path": "model.layers.0.self_attn.k_proj",
|
| 97 |
+
"in_features": 4096,
|
| 98 |
+
"out_features": 4096,
|
| 99 |
+
"rank": 512,
|
| 100 |
+
"sparse_nnz": 0
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"module_path": "model.layers.0.self_attn.v_proj",
|
| 104 |
+
"in_features": 4096,
|
| 105 |
+
"out_features": 4096,
|
| 106 |
+
"rank": 512,
|
| 107 |
+
"sparse_nnz": 0
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"module_path": "model.layers.0.self_attn.o_proj",
|
| 111 |
+
"in_features": 4096,
|
| 112 |
+
"out_features": 4096,
|
| 113 |
+
"rank": 512,
|
| 114 |
+
"sparse_nnz": 0
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"module_path": "model.layers.0.mlp.gate_proj",
|
| 118 |
+
"in_features": 4096,
|
| 119 |
+
"out_features": 11008,
|
| 120 |
+
"rank": 2132,
|
| 121 |
+
"sparse_nnz": 0
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"module_path": "model.layers.0.mlp.up_proj",
|
| 125 |
+
"in_features": 4096,
|
| 126 |
+
"out_features": 11008,
|
| 127 |
+
"rank": 2132,
|
| 128 |
+
"sparse_nnz": 0
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"module_path": "model.layers.0.mlp.down_proj",
|
| 132 |
+
"in_features": 11008,
|
| 133 |
+
"out_features": 4096,
|
| 134 |
+
"rank": 2132,
|
| 135 |
+
"sparse_nnz": 0
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"module_path": "model.layers.1.self_attn.q_proj",
|
| 139 |
+
"in_features": 4096,
|
| 140 |
+
"out_features": 4096,
|
| 141 |
+
"rank": 512,
|
| 142 |
+
"sparse_nnz": 0
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"module_path": "model.layers.1.self_attn.k_proj",
|
| 146 |
+
"in_features": 4096,
|
| 147 |
+
"out_features": 4096,
|
| 148 |
+
"rank": 512,
|
| 149 |
+
"sparse_nnz": 0
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"module_path": "model.layers.1.self_attn.v_proj",
|
| 153 |
+
"in_features": 4096,
|
| 154 |
+
"out_features": 4096,
|
| 155 |
+
"rank": 512,
|
| 156 |
+
"sparse_nnz": 0
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"module_path": "model.layers.1.self_attn.o_proj",
|
| 160 |
+
"in_features": 4096,
|
| 161 |
+
"out_features": 4096,
|
| 162 |
+
"rank": 512,
|
| 163 |
+
"sparse_nnz": 0
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"module_path": "model.layers.1.mlp.gate_proj",
|
| 167 |
+
"in_features": 4096,
|
| 168 |
+
"out_features": 11008,
|
| 169 |
+
"rank": 2132,
|
| 170 |
+
"sparse_nnz": 0
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"module_path": "model.layers.1.mlp.up_proj",
|
| 174 |
+
"in_features": 4096,
|
| 175 |
+
"out_features": 11008,
|
| 176 |
+
"rank": 2132,
|
| 177 |
+
"sparse_nnz": 0
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"module_path": "model.layers.1.mlp.down_proj",
|
| 181 |
+
"in_features": 11008,
|
| 182 |
+
"out_features": 4096,
|
| 183 |
+
"rank": 2132,
|
| 184 |
+
"sparse_nnz": 0
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"module_path": "model.layers.2.self_attn.q_proj",
|
| 188 |
+
"in_features": 4096,
|
| 189 |
+
"out_features": 4096,
|
| 190 |
+
"rank": 512,
|
| 191 |
+
"sparse_nnz": 0
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"module_path": "model.layers.2.self_attn.k_proj",
|
| 195 |
+
"in_features": 4096,
|
| 196 |
+
"out_features": 4096,
|
| 197 |
+
"rank": 512,
|
| 198 |
+
"sparse_nnz": 0
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"module_path": "model.layers.2.self_attn.v_proj",
|
| 202 |
+
"in_features": 4096,
|
| 203 |
+
"out_features": 4096,
|
| 204 |
+
"rank": 512,
|
| 205 |
+
"sparse_nnz": 0
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"module_path": "model.layers.2.self_attn.o_proj",
|
| 209 |
+
"in_features": 4096,
|
| 210 |
+
"out_features": 4096,
|
| 211 |
+
"rank": 512,
|
| 212 |
+
"sparse_nnz": 0
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"module_path": "model.layers.2.mlp.gate_proj",
|
| 216 |
+
"in_features": 4096,
|
| 217 |
+
"out_features": 11008,
|
| 218 |
+
"rank": 2132,
|
| 219 |
+
"sparse_nnz": 0
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"module_path": "model.layers.2.mlp.up_proj",
|
| 223 |
+
"in_features": 4096,
|
| 224 |
+
"out_features": 11008,
|
| 225 |
+
"rank": 2132,
|
| 226 |
+
"sparse_nnz": 0
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"module_path": "model.layers.2.mlp.down_proj",
|
| 230 |
+
"in_features": 11008,
|
| 231 |
+
"out_features": 4096,
|
| 232 |
+
"rank": 2132,
|
| 233 |
+
"sparse_nnz": 0
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"module_path": "model.layers.3.self_attn.q_proj",
|
| 237 |
+
"in_features": 4096,
|
| 238 |
+
"out_features": 4096,
|
| 239 |
+
"rank": 512,
|
| 240 |
+
"sparse_nnz": 0
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"module_path": "model.layers.3.self_attn.k_proj",
|
| 244 |
+
"in_features": 4096,
|
| 245 |
+
"out_features": 4096,
|
| 246 |
+
"rank": 512,
|
| 247 |
+
"sparse_nnz": 0
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"module_path": "model.layers.3.self_attn.v_proj",
|
| 251 |
+
"in_features": 4096,
|
| 252 |
+
"out_features": 4096,
|
| 253 |
+
"rank": 512,
|
| 254 |
+
"sparse_nnz": 0
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"module_path": "model.layers.3.self_attn.o_proj",
|
| 258 |
+
"in_features": 4096,
|
| 259 |
+
"out_features": 4096,
|
| 260 |
+
"rank": 512,
|
| 261 |
+
"sparse_nnz": 0
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"module_path": "model.layers.3.mlp.gate_proj",
|
| 265 |
+
"in_features": 4096,
|
| 266 |
+
"out_features": 11008,
|
| 267 |
+
"rank": 2132,
|
| 268 |
+
"sparse_nnz": 0
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"module_path": "model.layers.3.mlp.up_proj",
|
| 272 |
+
"in_features": 4096,
|
| 273 |
+
"out_features": 11008,
|
| 274 |
+
"rank": 2132,
|
| 275 |
+
"sparse_nnz": 0
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"module_path": "model.layers.3.mlp.down_proj",
|
| 279 |
+
"in_features": 11008,
|
| 280 |
+
"out_features": 4096,
|
| 281 |
+
"rank": 2132,
|
| 282 |
+
"sparse_nnz": 0
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"module_path": "model.layers.4.self_attn.q_proj",
|
| 286 |
+
"in_features": 4096,
|
| 287 |
+
"out_features": 4096,
|
| 288 |
+
"rank": 512,
|
| 289 |
+
"sparse_nnz": 0
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"module_path": "model.layers.4.self_attn.k_proj",
|
| 293 |
+
"in_features": 4096,
|
| 294 |
+
"out_features": 4096,
|
| 295 |
+
"rank": 512,
|
| 296 |
+
"sparse_nnz": 0
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"module_path": "model.layers.4.self_attn.v_proj",
|
| 300 |
+
"in_features": 4096,
|
| 301 |
+
"out_features": 4096,
|
| 302 |
+
"rank": 512,
|
| 303 |
+
"sparse_nnz": 0
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"module_path": "model.layers.4.self_attn.o_proj",
|
| 307 |
+
"in_features": 4096,
|
| 308 |
+
"out_features": 4096,
|
| 309 |
+
"rank": 512,
|
| 310 |
+
"sparse_nnz": 0
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"module_path": "model.layers.4.mlp.gate_proj",
|
| 314 |
+
"in_features": 4096,
|
| 315 |
+
"out_features": 11008,
|
| 316 |
+
"rank": 2132,
|
| 317 |
+
"sparse_nnz": 0
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"module_path": "model.layers.4.mlp.up_proj",
|
| 321 |
+
"in_features": 4096,
|
| 322 |
+
"out_features": 11008,
|
| 323 |
+
"rank": 2132,
|
| 324 |
+
"sparse_nnz": 0
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"module_path": "model.layers.4.mlp.down_proj",
|
| 328 |
+
"in_features": 11008,
|
| 329 |
+
"out_features": 4096,
|
| 330 |
+
"rank": 2132,
|
| 331 |
+
"sparse_nnz": 0
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"module_path": "model.layers.5.self_attn.q_proj",
|
| 335 |
+
"in_features": 4096,
|
| 336 |
+
"out_features": 4096,
|
| 337 |
+
"rank": 512,
|
| 338 |
+
"sparse_nnz": 0
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"module_path": "model.layers.5.self_attn.k_proj",
|
| 342 |
+
"in_features": 4096,
|
| 343 |
+
"out_features": 4096,
|
| 344 |
+
"rank": 512,
|
| 345 |
+
"sparse_nnz": 0
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"module_path": "model.layers.5.self_attn.v_proj",
|
| 349 |
+
"in_features": 4096,
|
| 350 |
+
"out_features": 4096,
|
| 351 |
+
"rank": 512,
|
| 352 |
+
"sparse_nnz": 0
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"module_path": "model.layers.5.self_attn.o_proj",
|
| 356 |
+
"in_features": 4096,
|
| 357 |
+
"out_features": 4096,
|
| 358 |
+
"rank": 512,
|
| 359 |
+
"sparse_nnz": 0
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"module_path": "model.layers.5.mlp.gate_proj",
|
| 363 |
+
"in_features": 4096,
|
| 364 |
+
"out_features": 11008,
|
| 365 |
+
"rank": 2132,
|
| 366 |
+
"sparse_nnz": 0
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"module_path": "model.layers.5.mlp.up_proj",
|
| 370 |
+
"in_features": 4096,
|
| 371 |
+
"out_features": 11008,
|
| 372 |
+
"rank": 2132,
|
| 373 |
+
"sparse_nnz": 0
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"module_path": "model.layers.5.mlp.down_proj",
|
| 377 |
+
"in_features": 11008,
|
| 378 |
+
"out_features": 4096,
|
| 379 |
+
"rank": 2132,
|
| 380 |
+
"sparse_nnz": 0
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"module_path": "model.layers.6.self_attn.q_proj",
|
| 384 |
+
"in_features": 4096,
|
| 385 |
+
"out_features": 4096,
|
| 386 |
+
"rank": 512,
|
| 387 |
+
"sparse_nnz": 0
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"module_path": "model.layers.6.self_attn.k_proj",
|
| 391 |
+
"in_features": 4096,
|
| 392 |
+
"out_features": 4096,
|
| 393 |
+
"rank": 512,
|
| 394 |
+
"sparse_nnz": 0
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"module_path": "model.layers.6.self_attn.v_proj",
|
| 398 |
+
"in_features": 4096,
|
| 399 |
+
"out_features": 4096,
|
| 400 |
+
"rank": 512,
|
| 401 |
+
"sparse_nnz": 0
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"module_path": "model.layers.6.self_attn.o_proj",
|
| 405 |
+
"in_features": 4096,
|
| 406 |
+
"out_features": 4096,
|
| 407 |
+
"rank": 512,
|
| 408 |
+
"sparse_nnz": 0
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"module_path": "model.layers.6.mlp.gate_proj",
|
| 412 |
+
"in_features": 4096,
|
| 413 |
+
"out_features": 11008,
|
| 414 |
+
"rank": 2132,
|
| 415 |
+
"sparse_nnz": 0
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"module_path": "model.layers.6.mlp.up_proj",
|
| 419 |
+
"in_features": 4096,
|
| 420 |
+
"out_features": 11008,
|
| 421 |
+
"rank": 2132,
|
| 422 |
+
"sparse_nnz": 0
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"module_path": "model.layers.6.mlp.down_proj",
|
| 426 |
+
"in_features": 11008,
|
| 427 |
+
"out_features": 4096,
|
| 428 |
+
"rank": 2132,
|
| 429 |
+
"sparse_nnz": 0
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"module_path": "model.layers.7.self_attn.q_proj",
|
| 433 |
+
"in_features": 4096,
|
| 434 |
+
"out_features": 4096,
|
| 435 |
+
"rank": 512,
|
| 436 |
+
"sparse_nnz": 0
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"module_path": "model.layers.7.self_attn.k_proj",
|
| 440 |
+
"in_features": 4096,
|
| 441 |
+
"out_features": 4096,
|
| 442 |
+
"rank": 512,
|
| 443 |
+
"sparse_nnz": 0
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"module_path": "model.layers.7.self_attn.v_proj",
|
| 447 |
+
"in_features": 4096,
|
| 448 |
+
"out_features": 4096,
|
| 449 |
+
"rank": 512,
|
| 450 |
+
"sparse_nnz": 0
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"module_path": "model.layers.7.self_attn.o_proj",
|
| 454 |
+
"in_features": 4096,
|
| 455 |
+
"out_features": 4096,
|
| 456 |
+
"rank": 512,
|
| 457 |
+
"sparse_nnz": 0
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"module_path": "model.layers.7.mlp.gate_proj",
|
| 461 |
+
"in_features": 4096,
|
| 462 |
+
"out_features": 11008,
|
| 463 |
+
"rank": 2132,
|
| 464 |
+
"sparse_nnz": 0
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"module_path": "model.layers.7.mlp.up_proj",
|
| 468 |
+
"in_features": 4096,
|
| 469 |
+
"out_features": 11008,
|
| 470 |
+
"rank": 2132,
|
| 471 |
+
"sparse_nnz": 0
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"module_path": "model.layers.7.mlp.down_proj",
|
| 475 |
+
"in_features": 11008,
|
| 476 |
+
"out_features": 4096,
|
| 477 |
+
"rank": 2132,
|
| 478 |
+
"sparse_nnz": 0
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"module_path": "model.layers.8.self_attn.q_proj",
|
| 482 |
+
"in_features": 4096,
|
| 483 |
+
"out_features": 4096,
|
| 484 |
+
"rank": 512,
|
| 485 |
+
"sparse_nnz": 0
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"module_path": "model.layers.8.self_attn.k_proj",
|
| 489 |
+
"in_features": 4096,
|
| 490 |
+
"out_features": 4096,
|
| 491 |
+
"rank": 512,
|
| 492 |
+
"sparse_nnz": 0
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"module_path": "model.layers.8.self_attn.v_proj",
|
| 496 |
+
"in_features": 4096,
|
| 497 |
+
"out_features": 4096,
|
| 498 |
+
"rank": 512,
|
| 499 |
+
"sparse_nnz": 0
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"module_path": "model.layers.8.self_attn.o_proj",
|
| 503 |
+
"in_features": 4096,
|
| 504 |
+
"out_features": 4096,
|
| 505 |
+
"rank": 512,
|
| 506 |
+
"sparse_nnz": 0
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"module_path": "model.layers.8.mlp.gate_proj",
|
| 510 |
+
"in_features": 4096,
|
| 511 |
+
"out_features": 11008,
|
| 512 |
+
"rank": 2132,
|
| 513 |
+
"sparse_nnz": 0
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"module_path": "model.layers.8.mlp.up_proj",
|
| 517 |
+
"in_features": 4096,
|
| 518 |
+
"out_features": 11008,
|
| 519 |
+
"rank": 2132,
|
| 520 |
+
"sparse_nnz": 0
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"module_path": "model.layers.8.mlp.down_proj",
|
| 524 |
+
"in_features": 11008,
|
| 525 |
+
"out_features": 4096,
|
| 526 |
+
"rank": 2132,
|
| 527 |
+
"sparse_nnz": 0
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"module_path": "model.layers.9.self_attn.q_proj",
|
| 531 |
+
"in_features": 4096,
|
| 532 |
+
"out_features": 4096,
|
| 533 |
+
"rank": 512,
|
| 534 |
+
"sparse_nnz": 0
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"module_path": "model.layers.9.self_attn.k_proj",
|
| 538 |
+
"in_features": 4096,
|
| 539 |
+
"out_features": 4096,
|
| 540 |
+
"rank": 512,
|
| 541 |
+
"sparse_nnz": 0
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"module_path": "model.layers.9.self_attn.v_proj",
|
| 545 |
+
"in_features": 4096,
|
| 546 |
+
"out_features": 4096,
|
| 547 |
+
"rank": 512,
|
| 548 |
+
"sparse_nnz": 0
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"module_path": "model.layers.9.self_attn.o_proj",
|
| 552 |
+
"in_features": 4096,
|
| 553 |
+
"out_features": 4096,
|
| 554 |
+
"rank": 512,
|
| 555 |
+
"sparse_nnz": 0
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"module_path": "model.layers.9.mlp.gate_proj",
|
| 559 |
+
"in_features": 4096,
|
| 560 |
+
"out_features": 11008,
|
| 561 |
+
"rank": 2132,
|
| 562 |
+
"sparse_nnz": 0
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"module_path": "model.layers.9.mlp.up_proj",
|
| 566 |
+
"in_features": 4096,
|
| 567 |
+
"out_features": 11008,
|
| 568 |
+
"rank": 2132,
|
| 569 |
+
"sparse_nnz": 0
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"module_path": "model.layers.9.mlp.down_proj",
|
| 573 |
+
"in_features": 11008,
|
| 574 |
+
"out_features": 4096,
|
| 575 |
+
"rank": 2132,
|
| 576 |
+
"sparse_nnz": 0
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"module_path": "model.layers.10.self_attn.q_proj",
|
| 580 |
+
"in_features": 4096,
|
| 581 |
+
"out_features": 4096,
|
| 582 |
+
"rank": 512,
|
| 583 |
+
"sparse_nnz": 0
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"module_path": "model.layers.10.self_attn.k_proj",
|
| 587 |
+
"in_features": 4096,
|
| 588 |
+
"out_features": 4096,
|
| 589 |
+
"rank": 512,
|
| 590 |
+
"sparse_nnz": 0
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"module_path": "model.layers.10.self_attn.v_proj",
|
| 594 |
+
"in_features": 4096,
|
| 595 |
+
"out_features": 4096,
|
| 596 |
+
"rank": 512,
|
| 597 |
+
"sparse_nnz": 0
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"module_path": "model.layers.10.self_attn.o_proj",
|
| 601 |
+
"in_features": 4096,
|
| 602 |
+
"out_features": 4096,
|
| 603 |
+
"rank": 512,
|
| 604 |
+
"sparse_nnz": 0
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"module_path": "model.layers.10.mlp.gate_proj",
|
| 608 |
+
"in_features": 4096,
|
| 609 |
+
"out_features": 11008,
|
| 610 |
+
"rank": 2132,
|
| 611 |
+
"sparse_nnz": 0
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"module_path": "model.layers.10.mlp.up_proj",
|
| 615 |
+
"in_features": 4096,
|
| 616 |
+
"out_features": 11008,
|
| 617 |
+
"rank": 2132,
|
| 618 |
+
"sparse_nnz": 0
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"module_path": "model.layers.10.mlp.down_proj",
|
| 622 |
+
"in_features": 11008,
|
| 623 |
+
"out_features": 4096,
|
| 624 |
+
"rank": 2132,
|
| 625 |
+
"sparse_nnz": 0
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"module_path": "model.layers.11.self_attn.q_proj",
|
| 629 |
+
"in_features": 4096,
|
| 630 |
+
"out_features": 4096,
|
| 631 |
+
"rank": 512,
|
| 632 |
+
"sparse_nnz": 0
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"module_path": "model.layers.11.self_attn.k_proj",
|
| 636 |
+
"in_features": 4096,
|
| 637 |
+
"out_features": 4096,
|
| 638 |
+
"rank": 512,
|
| 639 |
+
"sparse_nnz": 0
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"module_path": "model.layers.11.self_attn.v_proj",
|
| 643 |
+
"in_features": 4096,
|
| 644 |
+
"out_features": 4096,
|
| 645 |
+
"rank": 512,
|
| 646 |
+
"sparse_nnz": 0
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"module_path": "model.layers.11.self_attn.o_proj",
|
| 650 |
+
"in_features": 4096,
|
| 651 |
+
"out_features": 4096,
|
| 652 |
+
"rank": 512,
|
| 653 |
+
"sparse_nnz": 0
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"module_path": "model.layers.11.mlp.gate_proj",
|
| 657 |
+
"in_features": 4096,
|
| 658 |
+
"out_features": 11008,
|
| 659 |
+
"rank": 2132,
|
| 660 |
+
"sparse_nnz": 0
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"module_path": "model.layers.11.mlp.up_proj",
|
| 664 |
+
"in_features": 4096,
|
| 665 |
+
"out_features": 11008,
|
| 666 |
+
"rank": 2132,
|
| 667 |
+
"sparse_nnz": 0
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"module_path": "model.layers.11.mlp.down_proj",
|
| 671 |
+
"in_features": 11008,
|
| 672 |
+
"out_features": 4096,
|
| 673 |
+
"rank": 2132,
|
| 674 |
+
"sparse_nnz": 0
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"module_path": "model.layers.12.self_attn.q_proj",
|
| 678 |
+
"in_features": 4096,
|
| 679 |
+
"out_features": 4096,
|
| 680 |
+
"rank": 512,
|
| 681 |
+
"sparse_nnz": 0
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"module_path": "model.layers.12.self_attn.k_proj",
|
| 685 |
+
"in_features": 4096,
|
| 686 |
+
"out_features": 4096,
|
| 687 |
+
"rank": 512,
|
| 688 |
+
"sparse_nnz": 0
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"module_path": "model.layers.12.self_attn.v_proj",
|
| 692 |
+
"in_features": 4096,
|
| 693 |
+
"out_features": 4096,
|
| 694 |
+
"rank": 512,
|
| 695 |
+
"sparse_nnz": 0
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"module_path": "model.layers.12.self_attn.o_proj",
|
| 699 |
+
"in_features": 4096,
|
| 700 |
+
"out_features": 4096,
|
| 701 |
+
"rank": 512,
|
| 702 |
+
"sparse_nnz": 0
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"module_path": "model.layers.12.mlp.gate_proj",
|
| 706 |
+
"in_features": 4096,
|
| 707 |
+
"out_features": 11008,
|
| 708 |
+
"rank": 2132,
|
| 709 |
+
"sparse_nnz": 0
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"module_path": "model.layers.12.mlp.up_proj",
|
| 713 |
+
"in_features": 4096,
|
| 714 |
+
"out_features": 11008,
|
| 715 |
+
"rank": 2132,
|
| 716 |
+
"sparse_nnz": 0
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"module_path": "model.layers.12.mlp.down_proj",
|
| 720 |
+
"in_features": 11008,
|
| 721 |
+
"out_features": 4096,
|
| 722 |
+
"rank": 2132,
|
| 723 |
+
"sparse_nnz": 0
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"module_path": "model.layers.13.self_attn.q_proj",
|
| 727 |
+
"in_features": 4096,
|
| 728 |
+
"out_features": 4096,
|
| 729 |
+
"rank": 512,
|
| 730 |
+
"sparse_nnz": 0
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"module_path": "model.layers.13.self_attn.k_proj",
|
| 734 |
+
"in_features": 4096,
|
| 735 |
+
"out_features": 4096,
|
| 736 |
+
"rank": 512,
|
| 737 |
+
"sparse_nnz": 0
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"module_path": "model.layers.13.self_attn.v_proj",
|
| 741 |
+
"in_features": 4096,
|
| 742 |
+
"out_features": 4096,
|
| 743 |
+
"rank": 512,
|
| 744 |
+
"sparse_nnz": 0
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"module_path": "model.layers.13.self_attn.o_proj",
|
| 748 |
+
"in_features": 4096,
|
| 749 |
+
"out_features": 4096,
|
| 750 |
+
"rank": 512,
|
| 751 |
+
"sparse_nnz": 0
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"module_path": "model.layers.13.mlp.gate_proj",
|
| 755 |
+
"in_features": 4096,
|
| 756 |
+
"out_features": 11008,
|
| 757 |
+
"rank": 2132,
|
| 758 |
+
"sparse_nnz": 0
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"module_path": "model.layers.13.mlp.up_proj",
|
| 762 |
+
"in_features": 4096,
|
| 763 |
+
"out_features": 11008,
|
| 764 |
+
"rank": 2132,
|
| 765 |
+
"sparse_nnz": 0
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"module_path": "model.layers.13.mlp.down_proj",
|
| 769 |
+
"in_features": 11008,
|
| 770 |
+
"out_features": 4096,
|
| 771 |
+
"rank": 2132,
|
| 772 |
+
"sparse_nnz": 0
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"module_path": "model.layers.14.self_attn.q_proj",
|
| 776 |
+
"in_features": 4096,
|
| 777 |
+
"out_features": 4096,
|
| 778 |
+
"rank": 512,
|
| 779 |
+
"sparse_nnz": 0
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"module_path": "model.layers.14.self_attn.k_proj",
|
| 783 |
+
"in_features": 4096,
|
| 784 |
+
"out_features": 4096,
|
| 785 |
+
"rank": 512,
|
| 786 |
+
"sparse_nnz": 0
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"module_path": "model.layers.14.self_attn.v_proj",
|
| 790 |
+
"in_features": 4096,
|
| 791 |
+
"out_features": 4096,
|
| 792 |
+
"rank": 512,
|
| 793 |
+
"sparse_nnz": 0
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"module_path": "model.layers.14.self_attn.o_proj",
|
| 797 |
+
"in_features": 4096,
|
| 798 |
+
"out_features": 4096,
|
| 799 |
+
"rank": 512,
|
| 800 |
+
"sparse_nnz": 0
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"module_path": "model.layers.14.mlp.gate_proj",
|
| 804 |
+
"in_features": 4096,
|
| 805 |
+
"out_features": 11008,
|
| 806 |
+
"rank": 2132,
|
| 807 |
+
"sparse_nnz": 0
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"module_path": "model.layers.14.mlp.up_proj",
|
| 811 |
+
"in_features": 4096,
|
| 812 |
+
"out_features": 11008,
|
| 813 |
+
"rank": 2132,
|
| 814 |
+
"sparse_nnz": 0
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"module_path": "model.layers.14.mlp.down_proj",
|
| 818 |
+
"in_features": 11008,
|
| 819 |
+
"out_features": 4096,
|
| 820 |
+
"rank": 2132,
|
| 821 |
+
"sparse_nnz": 0
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"module_path": "model.layers.15.self_attn.q_proj",
|
| 825 |
+
"in_features": 4096,
|
| 826 |
+
"out_features": 4096,
|
| 827 |
+
"rank": 512,
|
| 828 |
+
"sparse_nnz": 0
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"module_path": "model.layers.15.self_attn.k_proj",
|
| 832 |
+
"in_features": 4096,
|
| 833 |
+
"out_features": 4096,
|
| 834 |
+
"rank": 512,
|
| 835 |
+
"sparse_nnz": 0
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"module_path": "model.layers.15.self_attn.v_proj",
|
| 839 |
+
"in_features": 4096,
|
| 840 |
+
"out_features": 4096,
|
| 841 |
+
"rank": 512,
|
| 842 |
+
"sparse_nnz": 0
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"module_path": "model.layers.15.self_attn.o_proj",
|
| 846 |
+
"in_features": 4096,
|
| 847 |
+
"out_features": 4096,
|
| 848 |
+
"rank": 512,
|
| 849 |
+
"sparse_nnz": 0
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"module_path": "model.layers.15.mlp.gate_proj",
|
| 853 |
+
"in_features": 4096,
|
| 854 |
+
"out_features": 11008,
|
| 855 |
+
"rank": 2132,
|
| 856 |
+
"sparse_nnz": 0
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"module_path": "model.layers.15.mlp.up_proj",
|
| 860 |
+
"in_features": 4096,
|
| 861 |
+
"out_features": 11008,
|
| 862 |
+
"rank": 2132,
|
| 863 |
+
"sparse_nnz": 0
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"module_path": "model.layers.15.mlp.down_proj",
|
| 867 |
+
"in_features": 11008,
|
| 868 |
+
"out_features": 4096,
|
| 869 |
+
"rank": 2132,
|
| 870 |
+
"sparse_nnz": 0
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"module_path": "model.layers.16.self_attn.q_proj",
|
| 874 |
+
"in_features": 4096,
|
| 875 |
+
"out_features": 4096,
|
| 876 |
+
"rank": 512,
|
| 877 |
+
"sparse_nnz": 0
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"module_path": "model.layers.16.self_attn.k_proj",
|
| 881 |
+
"in_features": 4096,
|
| 882 |
+
"out_features": 4096,
|
| 883 |
+
"rank": 512,
|
| 884 |
+
"sparse_nnz": 0
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"module_path": "model.layers.16.self_attn.v_proj",
|
| 888 |
+
"in_features": 4096,
|
| 889 |
+
"out_features": 4096,
|
| 890 |
+
"rank": 512,
|
| 891 |
+
"sparse_nnz": 0
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"module_path": "model.layers.16.self_attn.o_proj",
|
| 895 |
+
"in_features": 4096,
|
| 896 |
+
"out_features": 4096,
|
| 897 |
+
"rank": 512,
|
| 898 |
+
"sparse_nnz": 0
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"module_path": "model.layers.16.mlp.gate_proj",
|
| 902 |
+
"in_features": 4096,
|
| 903 |
+
"out_features": 11008,
|
| 904 |
+
"rank": 2132,
|
| 905 |
+
"sparse_nnz": 0
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"module_path": "model.layers.16.mlp.up_proj",
|
| 909 |
+
"in_features": 4096,
|
| 910 |
+
"out_features": 11008,
|
| 911 |
+
"rank": 2132,
|
| 912 |
+
"sparse_nnz": 0
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"module_path": "model.layers.16.mlp.down_proj",
|
| 916 |
+
"in_features": 11008,
|
| 917 |
+
"out_features": 4096,
|
| 918 |
+
"rank": 2132,
|
| 919 |
+
"sparse_nnz": 0
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"module_path": "model.layers.17.self_attn.q_proj",
|
| 923 |
+
"in_features": 4096,
|
| 924 |
+
"out_features": 4096,
|
| 925 |
+
"rank": 512,
|
| 926 |
+
"sparse_nnz": 0
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"module_path": "model.layers.17.self_attn.k_proj",
|
| 930 |
+
"in_features": 4096,
|
| 931 |
+
"out_features": 4096,
|
| 932 |
+
"rank": 512,
|
| 933 |
+
"sparse_nnz": 0
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"module_path": "model.layers.17.self_attn.v_proj",
|
| 937 |
+
"in_features": 4096,
|
| 938 |
+
"out_features": 4096,
|
| 939 |
+
"rank": 512,
|
| 940 |
+
"sparse_nnz": 0
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"module_path": "model.layers.17.self_attn.o_proj",
|
| 944 |
+
"in_features": 4096,
|
| 945 |
+
"out_features": 4096,
|
| 946 |
+
"rank": 512,
|
| 947 |
+
"sparse_nnz": 0
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"module_path": "model.layers.17.mlp.gate_proj",
|
| 951 |
+
"in_features": 4096,
|
| 952 |
+
"out_features": 11008,
|
| 953 |
+
"rank": 2132,
|
| 954 |
+
"sparse_nnz": 0
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"module_path": "model.layers.17.mlp.up_proj",
|
| 958 |
+
"in_features": 4096,
|
| 959 |
+
"out_features": 11008,
|
| 960 |
+
"rank": 2132,
|
| 961 |
+
"sparse_nnz": 0
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"module_path": "model.layers.17.mlp.down_proj",
|
| 965 |
+
"in_features": 11008,
|
| 966 |
+
"out_features": 4096,
|
| 967 |
+
"rank": 2132,
|
| 968 |
+
"sparse_nnz": 0
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"module_path": "model.layers.18.self_attn.q_proj",
|
| 972 |
+
"in_features": 4096,
|
| 973 |
+
"out_features": 4096,
|
| 974 |
+
"rank": 512,
|
| 975 |
+
"sparse_nnz": 0
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"module_path": "model.layers.18.self_attn.k_proj",
|
| 979 |
+
"in_features": 4096,
|
| 980 |
+
"out_features": 4096,
|
| 981 |
+
"rank": 512,
|
| 982 |
+
"sparse_nnz": 0
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"module_path": "model.layers.18.self_attn.v_proj",
|
| 986 |
+
"in_features": 4096,
|
| 987 |
+
"out_features": 4096,
|
| 988 |
+
"rank": 512,
|
| 989 |
+
"sparse_nnz": 0
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"module_path": "model.layers.18.self_attn.o_proj",
|
| 993 |
+
"in_features": 4096,
|
| 994 |
+
"out_features": 4096,
|
| 995 |
+
"rank": 512,
|
| 996 |
+
"sparse_nnz": 0
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"module_path": "model.layers.18.mlp.gate_proj",
|
| 1000 |
+
"in_features": 4096,
|
| 1001 |
+
"out_features": 11008,
|
| 1002 |
+
"rank": 2132,
|
| 1003 |
+
"sparse_nnz": 0
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"module_path": "model.layers.18.mlp.up_proj",
|
| 1007 |
+
"in_features": 4096,
|
| 1008 |
+
"out_features": 11008,
|
| 1009 |
+
"rank": 2132,
|
| 1010 |
+
"sparse_nnz": 0
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"module_path": "model.layers.18.mlp.down_proj",
|
| 1014 |
+
"in_features": 11008,
|
| 1015 |
+
"out_features": 4096,
|
| 1016 |
+
"rank": 2132,
|
| 1017 |
+
"sparse_nnz": 0
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"module_path": "model.layers.19.self_attn.q_proj",
|
| 1021 |
+
"in_features": 4096,
|
| 1022 |
+
"out_features": 4096,
|
| 1023 |
+
"rank": 512,
|
| 1024 |
+
"sparse_nnz": 0
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"module_path": "model.layers.19.self_attn.k_proj",
|
| 1028 |
+
"in_features": 4096,
|
| 1029 |
+
"out_features": 4096,
|
| 1030 |
+
"rank": 512,
|
| 1031 |
+
"sparse_nnz": 0
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"module_path": "model.layers.19.self_attn.v_proj",
|
| 1035 |
+
"in_features": 4096,
|
| 1036 |
+
"out_features": 4096,
|
| 1037 |
+
"rank": 512,
|
| 1038 |
+
"sparse_nnz": 0
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"module_path": "model.layers.19.self_attn.o_proj",
|
| 1042 |
+
"in_features": 4096,
|
| 1043 |
+
"out_features": 4096,
|
| 1044 |
+
"rank": 512,
|
| 1045 |
+
"sparse_nnz": 0
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"module_path": "model.layers.19.mlp.gate_proj",
|
| 1049 |
+
"in_features": 4096,
|
| 1050 |
+
"out_features": 11008,
|
| 1051 |
+
"rank": 2132,
|
| 1052 |
+
"sparse_nnz": 0
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"module_path": "model.layers.19.mlp.up_proj",
|
| 1056 |
+
"in_features": 4096,
|
| 1057 |
+
"out_features": 11008,
|
| 1058 |
+
"rank": 2132,
|
| 1059 |
+
"sparse_nnz": 0
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"module_path": "model.layers.19.mlp.down_proj",
|
| 1063 |
+
"in_features": 11008,
|
| 1064 |
+
"out_features": 4096,
|
| 1065 |
+
"rank": 2132,
|
| 1066 |
+
"sparse_nnz": 0
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"module_path": "model.layers.20.self_attn.q_proj",
|
| 1070 |
+
"in_features": 4096,
|
| 1071 |
+
"out_features": 4096,
|
| 1072 |
+
"rank": 512,
|
| 1073 |
+
"sparse_nnz": 0
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"module_path": "model.layers.20.self_attn.k_proj",
|
| 1077 |
+
"in_features": 4096,
|
| 1078 |
+
"out_features": 4096,
|
| 1079 |
+
"rank": 512,
|
| 1080 |
+
"sparse_nnz": 0
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"module_path": "model.layers.20.self_attn.v_proj",
|
| 1084 |
+
"in_features": 4096,
|
| 1085 |
+
"out_features": 4096,
|
| 1086 |
+
"rank": 512,
|
| 1087 |
+
"sparse_nnz": 0
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"module_path": "model.layers.20.self_attn.o_proj",
|
| 1091 |
+
"in_features": 4096,
|
| 1092 |
+
"out_features": 4096,
|
| 1093 |
+
"rank": 512,
|
| 1094 |
+
"sparse_nnz": 0
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"module_path": "model.layers.20.mlp.gate_proj",
|
| 1098 |
+
"in_features": 4096,
|
| 1099 |
+
"out_features": 11008,
|
| 1100 |
+
"rank": 2132,
|
| 1101 |
+
"sparse_nnz": 0
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"module_path": "model.layers.20.mlp.up_proj",
|
| 1105 |
+
"in_features": 4096,
|
| 1106 |
+
"out_features": 11008,
|
| 1107 |
+
"rank": 2132,
|
| 1108 |
+
"sparse_nnz": 0
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"module_path": "model.layers.20.mlp.down_proj",
|
| 1112 |
+
"in_features": 11008,
|
| 1113 |
+
"out_features": 4096,
|
| 1114 |
+
"rank": 2132,
|
| 1115 |
+
"sparse_nnz": 0
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"module_path": "model.layers.21.self_attn.q_proj",
|
| 1119 |
+
"in_features": 4096,
|
| 1120 |
+
"out_features": 4096,
|
| 1121 |
+
"rank": 512,
|
| 1122 |
+
"sparse_nnz": 0
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"module_path": "model.layers.21.self_attn.k_proj",
|
| 1126 |
+
"in_features": 4096,
|
| 1127 |
+
"out_features": 4096,
|
| 1128 |
+
"rank": 512,
|
| 1129 |
+
"sparse_nnz": 0
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"module_path": "model.layers.21.self_attn.v_proj",
|
| 1133 |
+
"in_features": 4096,
|
| 1134 |
+
"out_features": 4096,
|
| 1135 |
+
"rank": 512,
|
| 1136 |
+
"sparse_nnz": 0
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"module_path": "model.layers.21.self_attn.o_proj",
|
| 1140 |
+
"in_features": 4096,
|
| 1141 |
+
"out_features": 4096,
|
| 1142 |
+
"rank": 512,
|
| 1143 |
+
"sparse_nnz": 0
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"module_path": "model.layers.21.mlp.gate_proj",
|
| 1147 |
+
"in_features": 4096,
|
| 1148 |
+
"out_features": 11008,
|
| 1149 |
+
"rank": 2132,
|
| 1150 |
+
"sparse_nnz": 0
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"module_path": "model.layers.21.mlp.up_proj",
|
| 1154 |
+
"in_features": 4096,
|
| 1155 |
+
"out_features": 11008,
|
| 1156 |
+
"rank": 2132,
|
| 1157 |
+
"sparse_nnz": 0
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"module_path": "model.layers.21.mlp.down_proj",
|
| 1161 |
+
"in_features": 11008,
|
| 1162 |
+
"out_features": 4096,
|
| 1163 |
+
"rank": 2132,
|
| 1164 |
+
"sparse_nnz": 0
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"module_path": "model.layers.22.self_attn.q_proj",
|
| 1168 |
+
"in_features": 4096,
|
| 1169 |
+
"out_features": 4096,
|
| 1170 |
+
"rank": 512,
|
| 1171 |
+
"sparse_nnz": 0
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"module_path": "model.layers.22.self_attn.k_proj",
|
| 1175 |
+
"in_features": 4096,
|
| 1176 |
+
"out_features": 4096,
|
| 1177 |
+
"rank": 512,
|
| 1178 |
+
"sparse_nnz": 0
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"module_path": "model.layers.22.self_attn.v_proj",
|
| 1182 |
+
"in_features": 4096,
|
| 1183 |
+
"out_features": 4096,
|
| 1184 |
+
"rank": 512,
|
| 1185 |
+
"sparse_nnz": 0
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"module_path": "model.layers.22.self_attn.o_proj",
|
| 1189 |
+
"in_features": 4096,
|
| 1190 |
+
"out_features": 4096,
|
| 1191 |
+
"rank": 512,
|
| 1192 |
+
"sparse_nnz": 0
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"module_path": "model.layers.22.mlp.gate_proj",
|
| 1196 |
+
"in_features": 4096,
|
| 1197 |
+
"out_features": 11008,
|
| 1198 |
+
"rank": 2132,
|
| 1199 |
+
"sparse_nnz": 0
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"module_path": "model.layers.22.mlp.up_proj",
|
| 1203 |
+
"in_features": 4096,
|
| 1204 |
+
"out_features": 11008,
|
| 1205 |
+
"rank": 2132,
|
| 1206 |
+
"sparse_nnz": 0
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"module_path": "model.layers.22.mlp.down_proj",
|
| 1210 |
+
"in_features": 11008,
|
| 1211 |
+
"out_features": 4096,
|
| 1212 |
+
"rank": 2132,
|
| 1213 |
+
"sparse_nnz": 0
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"module_path": "model.layers.23.self_attn.q_proj",
|
| 1217 |
+
"in_features": 4096,
|
| 1218 |
+
"out_features": 4096,
|
| 1219 |
+
"rank": 512,
|
| 1220 |
+
"sparse_nnz": 0
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"module_path": "model.layers.23.self_attn.k_proj",
|
| 1224 |
+
"in_features": 4096,
|
| 1225 |
+
"out_features": 4096,
|
| 1226 |
+
"rank": 512,
|
| 1227 |
+
"sparse_nnz": 0
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"module_path": "model.layers.23.self_attn.v_proj",
|
| 1231 |
+
"in_features": 4096,
|
| 1232 |
+
"out_features": 4096,
|
| 1233 |
+
"rank": 512,
|
| 1234 |
+
"sparse_nnz": 0
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"module_path": "model.layers.23.self_attn.o_proj",
|
| 1238 |
+
"in_features": 4096,
|
| 1239 |
+
"out_features": 4096,
|
| 1240 |
+
"rank": 512,
|
| 1241 |
+
"sparse_nnz": 0
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"module_path": "model.layers.23.mlp.gate_proj",
|
| 1245 |
+
"in_features": 4096,
|
| 1246 |
+
"out_features": 11008,
|
| 1247 |
+
"rank": 2132,
|
| 1248 |
+
"sparse_nnz": 0
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"module_path": "model.layers.23.mlp.up_proj",
|
| 1252 |
+
"in_features": 4096,
|
| 1253 |
+
"out_features": 11008,
|
| 1254 |
+
"rank": 2132,
|
| 1255 |
+
"sparse_nnz": 0
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"module_path": "model.layers.23.mlp.down_proj",
|
| 1259 |
+
"in_features": 11008,
|
| 1260 |
+
"out_features": 4096,
|
| 1261 |
+
"rank": 2132,
|
| 1262 |
+
"sparse_nnz": 0
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"module_path": "model.layers.24.self_attn.q_proj",
|
| 1266 |
+
"in_features": 4096,
|
| 1267 |
+
"out_features": 4096,
|
| 1268 |
+
"rank": 512,
|
| 1269 |
+
"sparse_nnz": 0
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"module_path": "model.layers.24.self_attn.k_proj",
|
| 1273 |
+
"in_features": 4096,
|
| 1274 |
+
"out_features": 4096,
|
| 1275 |
+
"rank": 512,
|
| 1276 |
+
"sparse_nnz": 0
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"module_path": "model.layers.24.self_attn.v_proj",
|
| 1280 |
+
"in_features": 4096,
|
| 1281 |
+
"out_features": 4096,
|
| 1282 |
+
"rank": 512,
|
| 1283 |
+
"sparse_nnz": 0
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"module_path": "model.layers.24.self_attn.o_proj",
|
| 1287 |
+
"in_features": 4096,
|
| 1288 |
+
"out_features": 4096,
|
| 1289 |
+
"rank": 512,
|
| 1290 |
+
"sparse_nnz": 0
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"module_path": "model.layers.24.mlp.gate_proj",
|
| 1294 |
+
"in_features": 4096,
|
| 1295 |
+
"out_features": 11008,
|
| 1296 |
+
"rank": 2132,
|
| 1297 |
+
"sparse_nnz": 0
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"module_path": "model.layers.24.mlp.up_proj",
|
| 1301 |
+
"in_features": 4096,
|
| 1302 |
+
"out_features": 11008,
|
| 1303 |
+
"rank": 2132,
|
| 1304 |
+
"sparse_nnz": 0
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"module_path": "model.layers.24.mlp.down_proj",
|
| 1308 |
+
"in_features": 11008,
|
| 1309 |
+
"out_features": 4096,
|
| 1310 |
+
"rank": 2132,
|
| 1311 |
+
"sparse_nnz": 0
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"module_path": "model.layers.25.self_attn.q_proj",
|
| 1315 |
+
"in_features": 4096,
|
| 1316 |
+
"out_features": 4096,
|
| 1317 |
+
"rank": 512,
|
| 1318 |
+
"sparse_nnz": 0
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"module_path": "model.layers.25.self_attn.k_proj",
|
| 1322 |
+
"in_features": 4096,
|
| 1323 |
+
"out_features": 4096,
|
| 1324 |
+
"rank": 512,
|
| 1325 |
+
"sparse_nnz": 0
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"module_path": "model.layers.25.self_attn.v_proj",
|
| 1329 |
+
"in_features": 4096,
|
| 1330 |
+
"out_features": 4096,
|
| 1331 |
+
"rank": 512,
|
| 1332 |
+
"sparse_nnz": 0
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"module_path": "model.layers.25.self_attn.o_proj",
|
| 1336 |
+
"in_features": 4096,
|
| 1337 |
+
"out_features": 4096,
|
| 1338 |
+
"rank": 512,
|
| 1339 |
+
"sparse_nnz": 0
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"module_path": "model.layers.25.mlp.gate_proj",
|
| 1343 |
+
"in_features": 4096,
|
| 1344 |
+
"out_features": 11008,
|
| 1345 |
+
"rank": 2132,
|
| 1346 |
+
"sparse_nnz": 0
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"module_path": "model.layers.25.mlp.up_proj",
|
| 1350 |
+
"in_features": 4096,
|
| 1351 |
+
"out_features": 11008,
|
| 1352 |
+
"rank": 2132,
|
| 1353 |
+
"sparse_nnz": 0
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"module_path": "model.layers.25.mlp.down_proj",
|
| 1357 |
+
"in_features": 11008,
|
| 1358 |
+
"out_features": 4096,
|
| 1359 |
+
"rank": 2132,
|
| 1360 |
+
"sparse_nnz": 0
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"module_path": "model.layers.26.self_attn.q_proj",
|
| 1364 |
+
"in_features": 4096,
|
| 1365 |
+
"out_features": 4096,
|
| 1366 |
+
"rank": 512,
|
| 1367 |
+
"sparse_nnz": 0
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"module_path": "model.layers.26.self_attn.k_proj",
|
| 1371 |
+
"in_features": 4096,
|
| 1372 |
+
"out_features": 4096,
|
| 1373 |
+
"rank": 512,
|
| 1374 |
+
"sparse_nnz": 0
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"module_path": "model.layers.26.self_attn.v_proj",
|
| 1378 |
+
"in_features": 4096,
|
| 1379 |
+
"out_features": 4096,
|
| 1380 |
+
"rank": 512,
|
| 1381 |
+
"sparse_nnz": 0
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"module_path": "model.layers.26.self_attn.o_proj",
|
| 1385 |
+
"in_features": 4096,
|
| 1386 |
+
"out_features": 4096,
|
| 1387 |
+
"rank": 512,
|
| 1388 |
+
"sparse_nnz": 0
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"module_path": "model.layers.26.mlp.gate_proj",
|
| 1392 |
+
"in_features": 4096,
|
| 1393 |
+
"out_features": 11008,
|
| 1394 |
+
"rank": 2132,
|
| 1395 |
+
"sparse_nnz": 0
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"module_path": "model.layers.26.mlp.up_proj",
|
| 1399 |
+
"in_features": 4096,
|
| 1400 |
+
"out_features": 11008,
|
| 1401 |
+
"rank": 2132,
|
| 1402 |
+
"sparse_nnz": 0
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"module_path": "model.layers.26.mlp.down_proj",
|
| 1406 |
+
"in_features": 11008,
|
| 1407 |
+
"out_features": 4096,
|
| 1408 |
+
"rank": 2132,
|
| 1409 |
+
"sparse_nnz": 0
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"module_path": "model.layers.27.self_attn.q_proj",
|
| 1413 |
+
"in_features": 4096,
|
| 1414 |
+
"out_features": 4096,
|
| 1415 |
+
"rank": 512,
|
| 1416 |
+
"sparse_nnz": 0
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"module_path": "model.layers.27.self_attn.k_proj",
|
| 1420 |
+
"in_features": 4096,
|
| 1421 |
+
"out_features": 4096,
|
| 1422 |
+
"rank": 512,
|
| 1423 |
+
"sparse_nnz": 0
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"module_path": "model.layers.27.self_attn.v_proj",
|
| 1427 |
+
"in_features": 4096,
|
| 1428 |
+
"out_features": 4096,
|
| 1429 |
+
"rank": 512,
|
| 1430 |
+
"sparse_nnz": 0
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"module_path": "model.layers.27.self_attn.o_proj",
|
| 1434 |
+
"in_features": 4096,
|
| 1435 |
+
"out_features": 4096,
|
| 1436 |
+
"rank": 512,
|
| 1437 |
+
"sparse_nnz": 0
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"module_path": "model.layers.27.mlp.gate_proj",
|
| 1441 |
+
"in_features": 4096,
|
| 1442 |
+
"out_features": 11008,
|
| 1443 |
+
"rank": 2132,
|
| 1444 |
+
"sparse_nnz": 0
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"module_path": "model.layers.27.mlp.up_proj",
|
| 1448 |
+
"in_features": 4096,
|
| 1449 |
+
"out_features": 11008,
|
| 1450 |
+
"rank": 2132,
|
| 1451 |
+
"sparse_nnz": 0
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"module_path": "model.layers.27.mlp.down_proj",
|
| 1455 |
+
"in_features": 11008,
|
| 1456 |
+
"out_features": 4096,
|
| 1457 |
+
"rank": 2132,
|
| 1458 |
+
"sparse_nnz": 0
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"module_path": "model.layers.28.self_attn.q_proj",
|
| 1462 |
+
"in_features": 4096,
|
| 1463 |
+
"out_features": 4096,
|
| 1464 |
+
"rank": 512,
|
| 1465 |
+
"sparse_nnz": 0
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"module_path": "model.layers.28.self_attn.k_proj",
|
| 1469 |
+
"in_features": 4096,
|
| 1470 |
+
"out_features": 4096,
|
| 1471 |
+
"rank": 512,
|
| 1472 |
+
"sparse_nnz": 0
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"module_path": "model.layers.28.self_attn.v_proj",
|
| 1476 |
+
"in_features": 4096,
|
| 1477 |
+
"out_features": 4096,
|
| 1478 |
+
"rank": 512,
|
| 1479 |
+
"sparse_nnz": 0
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"module_path": "model.layers.28.self_attn.o_proj",
|
| 1483 |
+
"in_features": 4096,
|
| 1484 |
+
"out_features": 4096,
|
| 1485 |
+
"rank": 512,
|
| 1486 |
+
"sparse_nnz": 0
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"module_path": "model.layers.28.mlp.gate_proj",
|
| 1490 |
+
"in_features": 4096,
|
| 1491 |
+
"out_features": 11008,
|
| 1492 |
+
"rank": 2132,
|
| 1493 |
+
"sparse_nnz": 0
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"module_path": "model.layers.28.mlp.up_proj",
|
| 1497 |
+
"in_features": 4096,
|
| 1498 |
+
"out_features": 11008,
|
| 1499 |
+
"rank": 2132,
|
| 1500 |
+
"sparse_nnz": 0
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"module_path": "model.layers.28.mlp.down_proj",
|
| 1504 |
+
"in_features": 11008,
|
| 1505 |
+
"out_features": 4096,
|
| 1506 |
+
"rank": 2132,
|
| 1507 |
+
"sparse_nnz": 0
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"module_path": "model.layers.29.self_attn.q_proj",
|
| 1511 |
+
"in_features": 4096,
|
| 1512 |
+
"out_features": 4096,
|
| 1513 |
+
"rank": 512,
|
| 1514 |
+
"sparse_nnz": 0
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"module_path": "model.layers.29.self_attn.k_proj",
|
| 1518 |
+
"in_features": 4096,
|
| 1519 |
+
"out_features": 4096,
|
| 1520 |
+
"rank": 512,
|
| 1521 |
+
"sparse_nnz": 0
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"module_path": "model.layers.29.self_attn.v_proj",
|
| 1525 |
+
"in_features": 4096,
|
| 1526 |
+
"out_features": 4096,
|
| 1527 |
+
"rank": 512,
|
| 1528 |
+
"sparse_nnz": 0
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"module_path": "model.layers.29.self_attn.o_proj",
|
| 1532 |
+
"in_features": 4096,
|
| 1533 |
+
"out_features": 4096,
|
| 1534 |
+
"rank": 512,
|
| 1535 |
+
"sparse_nnz": 0
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"module_path": "model.layers.29.mlp.gate_proj",
|
| 1539 |
+
"in_features": 4096,
|
| 1540 |
+
"out_features": 11008,
|
| 1541 |
+
"rank": 2132,
|
| 1542 |
+
"sparse_nnz": 0
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"module_path": "model.layers.29.mlp.up_proj",
|
| 1546 |
+
"in_features": 4096,
|
| 1547 |
+
"out_features": 11008,
|
| 1548 |
+
"rank": 2132,
|
| 1549 |
+
"sparse_nnz": 0
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"module_path": "model.layers.29.mlp.down_proj",
|
| 1553 |
+
"in_features": 11008,
|
| 1554 |
+
"out_features": 4096,
|
| 1555 |
+
"rank": 2132,
|
| 1556 |
+
"sparse_nnz": 0
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"module_path": "model.layers.30.self_attn.q_proj",
|
| 1560 |
+
"in_features": 4096,
|
| 1561 |
+
"out_features": 4096,
|
| 1562 |
+
"rank": 512,
|
| 1563 |
+
"sparse_nnz": 0
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"module_path": "model.layers.30.self_attn.k_proj",
|
| 1567 |
+
"in_features": 4096,
|
| 1568 |
+
"out_features": 4096,
|
| 1569 |
+
"rank": 512,
|
| 1570 |
+
"sparse_nnz": 0
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"module_path": "model.layers.30.self_attn.v_proj",
|
| 1574 |
+
"in_features": 4096,
|
| 1575 |
+
"out_features": 4096,
|
| 1576 |
+
"rank": 512,
|
| 1577 |
+
"sparse_nnz": 0
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"module_path": "model.layers.30.self_attn.o_proj",
|
| 1581 |
+
"in_features": 4096,
|
| 1582 |
+
"out_features": 4096,
|
| 1583 |
+
"rank": 512,
|
| 1584 |
+
"sparse_nnz": 0
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"module_path": "model.layers.30.mlp.gate_proj",
|
| 1588 |
+
"in_features": 4096,
|
| 1589 |
+
"out_features": 11008,
|
| 1590 |
+
"rank": 2132,
|
| 1591 |
+
"sparse_nnz": 0
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"module_path": "model.layers.30.mlp.up_proj",
|
| 1595 |
+
"in_features": 4096,
|
| 1596 |
+
"out_features": 11008,
|
| 1597 |
+
"rank": 2132,
|
| 1598 |
+
"sparse_nnz": 0
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"module_path": "model.layers.30.mlp.down_proj",
|
| 1602 |
+
"in_features": 11008,
|
| 1603 |
+
"out_features": 4096,
|
| 1604 |
+
"rank": 2132,
|
| 1605 |
+
"sparse_nnz": 0
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"module_path": "model.layers.31.self_attn.q_proj",
|
| 1609 |
+
"in_features": 4096,
|
| 1610 |
+
"out_features": 4096,
|
| 1611 |
+
"rank": 512,
|
| 1612 |
+
"sparse_nnz": 0
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"module_path": "model.layers.31.self_attn.k_proj",
|
| 1616 |
+
"in_features": 4096,
|
| 1617 |
+
"out_features": 4096,
|
| 1618 |
+
"rank": 512,
|
| 1619 |
+
"sparse_nnz": 0
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"module_path": "model.layers.31.self_attn.v_proj",
|
| 1623 |
+
"in_features": 4096,
|
| 1624 |
+
"out_features": 4096,
|
| 1625 |
+
"rank": 512,
|
| 1626 |
+
"sparse_nnz": 0
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"module_path": "model.layers.31.self_attn.o_proj",
|
| 1630 |
+
"in_features": 4096,
|
| 1631 |
+
"out_features": 4096,
|
| 1632 |
+
"rank": 512,
|
| 1633 |
+
"sparse_nnz": 0
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"module_path": "model.layers.31.mlp.gate_proj",
|
| 1637 |
+
"in_features": 4096,
|
| 1638 |
+
"out_features": 11008,
|
| 1639 |
+
"rank": 2132,
|
| 1640 |
+
"sparse_nnz": 0
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"module_path": "model.layers.31.mlp.up_proj",
|
| 1644 |
+
"in_features": 4096,
|
| 1645 |
+
"out_features": 11008,
|
| 1646 |
+
"rank": 2132,
|
| 1647 |
+
"sparse_nnz": 0
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"module_path": "model.layers.31.mlp.down_proj",
|
| 1651 |
+
"in_features": 11008,
|
| 1652 |
+
"out_features": 4096,
|
| 1653 |
+
"rank": 2132,
|
| 1654 |
+
"sparse_nnz": 0
|
| 1655 |
+
}
|
| 1656 |
+
]
|
| 1657 |
+
}
|
factorized_modules.json
ADDED
|
@@ -0,0 +1,1570 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"module_path": "model.layers.0.self_attn.q_proj",
|
| 4 |
+
"in_features": 4096,
|
| 5 |
+
"out_features": 4096,
|
| 6 |
+
"rank": 512,
|
| 7 |
+
"sparse_nnz": 0
|
| 8 |
+
},
|
| 9 |
+
{
|
| 10 |
+
"module_path": "model.layers.0.self_attn.k_proj",
|
| 11 |
+
"in_features": 4096,
|
| 12 |
+
"out_features": 4096,
|
| 13 |
+
"rank": 512,
|
| 14 |
+
"sparse_nnz": 0
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"module_path": "model.layers.0.self_attn.v_proj",
|
| 18 |
+
"in_features": 4096,
|
| 19 |
+
"out_features": 4096,
|
| 20 |
+
"rank": 512,
|
| 21 |
+
"sparse_nnz": 0
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"module_path": "model.layers.0.self_attn.o_proj",
|
| 25 |
+
"in_features": 4096,
|
| 26 |
+
"out_features": 4096,
|
| 27 |
+
"rank": 512,
|
| 28 |
+
"sparse_nnz": 0
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"module_path": "model.layers.0.mlp.gate_proj",
|
| 32 |
+
"in_features": 4096,
|
| 33 |
+
"out_features": 11008,
|
| 34 |
+
"rank": 2132,
|
| 35 |
+
"sparse_nnz": 0
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"module_path": "model.layers.0.mlp.up_proj",
|
| 39 |
+
"in_features": 4096,
|
| 40 |
+
"out_features": 11008,
|
| 41 |
+
"rank": 2132,
|
| 42 |
+
"sparse_nnz": 0
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"module_path": "model.layers.0.mlp.down_proj",
|
| 46 |
+
"in_features": 11008,
|
| 47 |
+
"out_features": 4096,
|
| 48 |
+
"rank": 2132,
|
| 49 |
+
"sparse_nnz": 0
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"module_path": "model.layers.1.self_attn.q_proj",
|
| 53 |
+
"in_features": 4096,
|
| 54 |
+
"out_features": 4096,
|
| 55 |
+
"rank": 512,
|
| 56 |
+
"sparse_nnz": 0
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"module_path": "model.layers.1.self_attn.k_proj",
|
| 60 |
+
"in_features": 4096,
|
| 61 |
+
"out_features": 4096,
|
| 62 |
+
"rank": 512,
|
| 63 |
+
"sparse_nnz": 0
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"module_path": "model.layers.1.self_attn.v_proj",
|
| 67 |
+
"in_features": 4096,
|
| 68 |
+
"out_features": 4096,
|
| 69 |
+
"rank": 512,
|
| 70 |
+
"sparse_nnz": 0
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"module_path": "model.layers.1.self_attn.o_proj",
|
| 74 |
+
"in_features": 4096,
|
| 75 |
+
"out_features": 4096,
|
| 76 |
+
"rank": 512,
|
| 77 |
+
"sparse_nnz": 0
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"module_path": "model.layers.1.mlp.gate_proj",
|
| 81 |
+
"in_features": 4096,
|
| 82 |
+
"out_features": 11008,
|
| 83 |
+
"rank": 2132,
|
| 84 |
+
"sparse_nnz": 0
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"module_path": "model.layers.1.mlp.up_proj",
|
| 88 |
+
"in_features": 4096,
|
| 89 |
+
"out_features": 11008,
|
| 90 |
+
"rank": 2132,
|
| 91 |
+
"sparse_nnz": 0
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"module_path": "model.layers.1.mlp.down_proj",
|
| 95 |
+
"in_features": 11008,
|
| 96 |
+
"out_features": 4096,
|
| 97 |
+
"rank": 2132,
|
| 98 |
+
"sparse_nnz": 0
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"module_path": "model.layers.2.self_attn.q_proj",
|
| 102 |
+
"in_features": 4096,
|
| 103 |
+
"out_features": 4096,
|
| 104 |
+
"rank": 512,
|
| 105 |
+
"sparse_nnz": 0
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"module_path": "model.layers.2.self_attn.k_proj",
|
| 109 |
+
"in_features": 4096,
|
| 110 |
+
"out_features": 4096,
|
| 111 |
+
"rank": 512,
|
| 112 |
+
"sparse_nnz": 0
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"module_path": "model.layers.2.self_attn.v_proj",
|
| 116 |
+
"in_features": 4096,
|
| 117 |
+
"out_features": 4096,
|
| 118 |
+
"rank": 512,
|
| 119 |
+
"sparse_nnz": 0
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"module_path": "model.layers.2.self_attn.o_proj",
|
| 123 |
+
"in_features": 4096,
|
| 124 |
+
"out_features": 4096,
|
| 125 |
+
"rank": 512,
|
| 126 |
+
"sparse_nnz": 0
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"module_path": "model.layers.2.mlp.gate_proj",
|
| 130 |
+
"in_features": 4096,
|
| 131 |
+
"out_features": 11008,
|
| 132 |
+
"rank": 2132,
|
| 133 |
+
"sparse_nnz": 0
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"module_path": "model.layers.2.mlp.up_proj",
|
| 137 |
+
"in_features": 4096,
|
| 138 |
+
"out_features": 11008,
|
| 139 |
+
"rank": 2132,
|
| 140 |
+
"sparse_nnz": 0
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"module_path": "model.layers.2.mlp.down_proj",
|
| 144 |
+
"in_features": 11008,
|
| 145 |
+
"out_features": 4096,
|
| 146 |
+
"rank": 2132,
|
| 147 |
+
"sparse_nnz": 0
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"module_path": "model.layers.3.self_attn.q_proj",
|
| 151 |
+
"in_features": 4096,
|
| 152 |
+
"out_features": 4096,
|
| 153 |
+
"rank": 512,
|
| 154 |
+
"sparse_nnz": 0
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"module_path": "model.layers.3.self_attn.k_proj",
|
| 158 |
+
"in_features": 4096,
|
| 159 |
+
"out_features": 4096,
|
| 160 |
+
"rank": 512,
|
| 161 |
+
"sparse_nnz": 0
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"module_path": "model.layers.3.self_attn.v_proj",
|
| 165 |
+
"in_features": 4096,
|
| 166 |
+
"out_features": 4096,
|
| 167 |
+
"rank": 512,
|
| 168 |
+
"sparse_nnz": 0
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"module_path": "model.layers.3.self_attn.o_proj",
|
| 172 |
+
"in_features": 4096,
|
| 173 |
+
"out_features": 4096,
|
| 174 |
+
"rank": 512,
|
| 175 |
+
"sparse_nnz": 0
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"module_path": "model.layers.3.mlp.gate_proj",
|
| 179 |
+
"in_features": 4096,
|
| 180 |
+
"out_features": 11008,
|
| 181 |
+
"rank": 2132,
|
| 182 |
+
"sparse_nnz": 0
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"module_path": "model.layers.3.mlp.up_proj",
|
| 186 |
+
"in_features": 4096,
|
| 187 |
+
"out_features": 11008,
|
| 188 |
+
"rank": 2132,
|
| 189 |
+
"sparse_nnz": 0
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"module_path": "model.layers.3.mlp.down_proj",
|
| 193 |
+
"in_features": 11008,
|
| 194 |
+
"out_features": 4096,
|
| 195 |
+
"rank": 2132,
|
| 196 |
+
"sparse_nnz": 0
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"module_path": "model.layers.4.self_attn.q_proj",
|
| 200 |
+
"in_features": 4096,
|
| 201 |
+
"out_features": 4096,
|
| 202 |
+
"rank": 512,
|
| 203 |
+
"sparse_nnz": 0
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"module_path": "model.layers.4.self_attn.k_proj",
|
| 207 |
+
"in_features": 4096,
|
| 208 |
+
"out_features": 4096,
|
| 209 |
+
"rank": 512,
|
| 210 |
+
"sparse_nnz": 0
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"module_path": "model.layers.4.self_attn.v_proj",
|
| 214 |
+
"in_features": 4096,
|
| 215 |
+
"out_features": 4096,
|
| 216 |
+
"rank": 512,
|
| 217 |
+
"sparse_nnz": 0
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"module_path": "model.layers.4.self_attn.o_proj",
|
| 221 |
+
"in_features": 4096,
|
| 222 |
+
"out_features": 4096,
|
| 223 |
+
"rank": 512,
|
| 224 |
+
"sparse_nnz": 0
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"module_path": "model.layers.4.mlp.gate_proj",
|
| 228 |
+
"in_features": 4096,
|
| 229 |
+
"out_features": 11008,
|
| 230 |
+
"rank": 2132,
|
| 231 |
+
"sparse_nnz": 0
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"module_path": "model.layers.4.mlp.up_proj",
|
| 235 |
+
"in_features": 4096,
|
| 236 |
+
"out_features": 11008,
|
| 237 |
+
"rank": 2132,
|
| 238 |
+
"sparse_nnz": 0
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"module_path": "model.layers.4.mlp.down_proj",
|
| 242 |
+
"in_features": 11008,
|
| 243 |
+
"out_features": 4096,
|
| 244 |
+
"rank": 2132,
|
| 245 |
+
"sparse_nnz": 0
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"module_path": "model.layers.5.self_attn.q_proj",
|
| 249 |
+
"in_features": 4096,
|
| 250 |
+
"out_features": 4096,
|
| 251 |
+
"rank": 512,
|
| 252 |
+
"sparse_nnz": 0
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"module_path": "model.layers.5.self_attn.k_proj",
|
| 256 |
+
"in_features": 4096,
|
| 257 |
+
"out_features": 4096,
|
| 258 |
+
"rank": 512,
|
| 259 |
+
"sparse_nnz": 0
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"module_path": "model.layers.5.self_attn.v_proj",
|
| 263 |
+
"in_features": 4096,
|
| 264 |
+
"out_features": 4096,
|
| 265 |
+
"rank": 512,
|
| 266 |
+
"sparse_nnz": 0
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"module_path": "model.layers.5.self_attn.o_proj",
|
| 270 |
+
"in_features": 4096,
|
| 271 |
+
"out_features": 4096,
|
| 272 |
+
"rank": 512,
|
| 273 |
+
"sparse_nnz": 0
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"module_path": "model.layers.5.mlp.gate_proj",
|
| 277 |
+
"in_features": 4096,
|
| 278 |
+
"out_features": 11008,
|
| 279 |
+
"rank": 2132,
|
| 280 |
+
"sparse_nnz": 0
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"module_path": "model.layers.5.mlp.up_proj",
|
| 284 |
+
"in_features": 4096,
|
| 285 |
+
"out_features": 11008,
|
| 286 |
+
"rank": 2132,
|
| 287 |
+
"sparse_nnz": 0
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"module_path": "model.layers.5.mlp.down_proj",
|
| 291 |
+
"in_features": 11008,
|
| 292 |
+
"out_features": 4096,
|
| 293 |
+
"rank": 2132,
|
| 294 |
+
"sparse_nnz": 0
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"module_path": "model.layers.6.self_attn.q_proj",
|
| 298 |
+
"in_features": 4096,
|
| 299 |
+
"out_features": 4096,
|
| 300 |
+
"rank": 512,
|
| 301 |
+
"sparse_nnz": 0
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"module_path": "model.layers.6.self_attn.k_proj",
|
| 305 |
+
"in_features": 4096,
|
| 306 |
+
"out_features": 4096,
|
| 307 |
+
"rank": 512,
|
| 308 |
+
"sparse_nnz": 0
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"module_path": "model.layers.6.self_attn.v_proj",
|
| 312 |
+
"in_features": 4096,
|
| 313 |
+
"out_features": 4096,
|
| 314 |
+
"rank": 512,
|
| 315 |
+
"sparse_nnz": 0
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"module_path": "model.layers.6.self_attn.o_proj",
|
| 319 |
+
"in_features": 4096,
|
| 320 |
+
"out_features": 4096,
|
| 321 |
+
"rank": 512,
|
| 322 |
+
"sparse_nnz": 0
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"module_path": "model.layers.6.mlp.gate_proj",
|
| 326 |
+
"in_features": 4096,
|
| 327 |
+
"out_features": 11008,
|
| 328 |
+
"rank": 2132,
|
| 329 |
+
"sparse_nnz": 0
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"module_path": "model.layers.6.mlp.up_proj",
|
| 333 |
+
"in_features": 4096,
|
| 334 |
+
"out_features": 11008,
|
| 335 |
+
"rank": 2132,
|
| 336 |
+
"sparse_nnz": 0
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"module_path": "model.layers.6.mlp.down_proj",
|
| 340 |
+
"in_features": 11008,
|
| 341 |
+
"out_features": 4096,
|
| 342 |
+
"rank": 2132,
|
| 343 |
+
"sparse_nnz": 0
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"module_path": "model.layers.7.self_attn.q_proj",
|
| 347 |
+
"in_features": 4096,
|
| 348 |
+
"out_features": 4096,
|
| 349 |
+
"rank": 512,
|
| 350 |
+
"sparse_nnz": 0
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"module_path": "model.layers.7.self_attn.k_proj",
|
| 354 |
+
"in_features": 4096,
|
| 355 |
+
"out_features": 4096,
|
| 356 |
+
"rank": 512,
|
| 357 |
+
"sparse_nnz": 0
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"module_path": "model.layers.7.self_attn.v_proj",
|
| 361 |
+
"in_features": 4096,
|
| 362 |
+
"out_features": 4096,
|
| 363 |
+
"rank": 512,
|
| 364 |
+
"sparse_nnz": 0
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"module_path": "model.layers.7.self_attn.o_proj",
|
| 368 |
+
"in_features": 4096,
|
| 369 |
+
"out_features": 4096,
|
| 370 |
+
"rank": 512,
|
| 371 |
+
"sparse_nnz": 0
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"module_path": "model.layers.7.mlp.gate_proj",
|
| 375 |
+
"in_features": 4096,
|
| 376 |
+
"out_features": 11008,
|
| 377 |
+
"rank": 2132,
|
| 378 |
+
"sparse_nnz": 0
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"module_path": "model.layers.7.mlp.up_proj",
|
| 382 |
+
"in_features": 4096,
|
| 383 |
+
"out_features": 11008,
|
| 384 |
+
"rank": 2132,
|
| 385 |
+
"sparse_nnz": 0
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"module_path": "model.layers.7.mlp.down_proj",
|
| 389 |
+
"in_features": 11008,
|
| 390 |
+
"out_features": 4096,
|
| 391 |
+
"rank": 2132,
|
| 392 |
+
"sparse_nnz": 0
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"module_path": "model.layers.8.self_attn.q_proj",
|
| 396 |
+
"in_features": 4096,
|
| 397 |
+
"out_features": 4096,
|
| 398 |
+
"rank": 512,
|
| 399 |
+
"sparse_nnz": 0
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"module_path": "model.layers.8.self_attn.k_proj",
|
| 403 |
+
"in_features": 4096,
|
| 404 |
+
"out_features": 4096,
|
| 405 |
+
"rank": 512,
|
| 406 |
+
"sparse_nnz": 0
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"module_path": "model.layers.8.self_attn.v_proj",
|
| 410 |
+
"in_features": 4096,
|
| 411 |
+
"out_features": 4096,
|
| 412 |
+
"rank": 512,
|
| 413 |
+
"sparse_nnz": 0
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"module_path": "model.layers.8.self_attn.o_proj",
|
| 417 |
+
"in_features": 4096,
|
| 418 |
+
"out_features": 4096,
|
| 419 |
+
"rank": 512,
|
| 420 |
+
"sparse_nnz": 0
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"module_path": "model.layers.8.mlp.gate_proj",
|
| 424 |
+
"in_features": 4096,
|
| 425 |
+
"out_features": 11008,
|
| 426 |
+
"rank": 2132,
|
| 427 |
+
"sparse_nnz": 0
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"module_path": "model.layers.8.mlp.up_proj",
|
| 431 |
+
"in_features": 4096,
|
| 432 |
+
"out_features": 11008,
|
| 433 |
+
"rank": 2132,
|
| 434 |
+
"sparse_nnz": 0
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"module_path": "model.layers.8.mlp.down_proj",
|
| 438 |
+
"in_features": 11008,
|
| 439 |
+
"out_features": 4096,
|
| 440 |
+
"rank": 2132,
|
| 441 |
+
"sparse_nnz": 0
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"module_path": "model.layers.9.self_attn.q_proj",
|
| 445 |
+
"in_features": 4096,
|
| 446 |
+
"out_features": 4096,
|
| 447 |
+
"rank": 512,
|
| 448 |
+
"sparse_nnz": 0
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"module_path": "model.layers.9.self_attn.k_proj",
|
| 452 |
+
"in_features": 4096,
|
| 453 |
+
"out_features": 4096,
|
| 454 |
+
"rank": 512,
|
| 455 |
+
"sparse_nnz": 0
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"module_path": "model.layers.9.self_attn.v_proj",
|
| 459 |
+
"in_features": 4096,
|
| 460 |
+
"out_features": 4096,
|
| 461 |
+
"rank": 512,
|
| 462 |
+
"sparse_nnz": 0
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"module_path": "model.layers.9.self_attn.o_proj",
|
| 466 |
+
"in_features": 4096,
|
| 467 |
+
"out_features": 4096,
|
| 468 |
+
"rank": 512,
|
| 469 |
+
"sparse_nnz": 0
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"module_path": "model.layers.9.mlp.gate_proj",
|
| 473 |
+
"in_features": 4096,
|
| 474 |
+
"out_features": 11008,
|
| 475 |
+
"rank": 2132,
|
| 476 |
+
"sparse_nnz": 0
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"module_path": "model.layers.9.mlp.up_proj",
|
| 480 |
+
"in_features": 4096,
|
| 481 |
+
"out_features": 11008,
|
| 482 |
+
"rank": 2132,
|
| 483 |
+
"sparse_nnz": 0
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"module_path": "model.layers.9.mlp.down_proj",
|
| 487 |
+
"in_features": 11008,
|
| 488 |
+
"out_features": 4096,
|
| 489 |
+
"rank": 2132,
|
| 490 |
+
"sparse_nnz": 0
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"module_path": "model.layers.10.self_attn.q_proj",
|
| 494 |
+
"in_features": 4096,
|
| 495 |
+
"out_features": 4096,
|
| 496 |
+
"rank": 512,
|
| 497 |
+
"sparse_nnz": 0
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"module_path": "model.layers.10.self_attn.k_proj",
|
| 501 |
+
"in_features": 4096,
|
| 502 |
+
"out_features": 4096,
|
| 503 |
+
"rank": 512,
|
| 504 |
+
"sparse_nnz": 0
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"module_path": "model.layers.10.self_attn.v_proj",
|
| 508 |
+
"in_features": 4096,
|
| 509 |
+
"out_features": 4096,
|
| 510 |
+
"rank": 512,
|
| 511 |
+
"sparse_nnz": 0
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"module_path": "model.layers.10.self_attn.o_proj",
|
| 515 |
+
"in_features": 4096,
|
| 516 |
+
"out_features": 4096,
|
| 517 |
+
"rank": 512,
|
| 518 |
+
"sparse_nnz": 0
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"module_path": "model.layers.10.mlp.gate_proj",
|
| 522 |
+
"in_features": 4096,
|
| 523 |
+
"out_features": 11008,
|
| 524 |
+
"rank": 2132,
|
| 525 |
+
"sparse_nnz": 0
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"module_path": "model.layers.10.mlp.up_proj",
|
| 529 |
+
"in_features": 4096,
|
| 530 |
+
"out_features": 11008,
|
| 531 |
+
"rank": 2132,
|
| 532 |
+
"sparse_nnz": 0
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"module_path": "model.layers.10.mlp.down_proj",
|
| 536 |
+
"in_features": 11008,
|
| 537 |
+
"out_features": 4096,
|
| 538 |
+
"rank": 2132,
|
| 539 |
+
"sparse_nnz": 0
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"module_path": "model.layers.11.self_attn.q_proj",
|
| 543 |
+
"in_features": 4096,
|
| 544 |
+
"out_features": 4096,
|
| 545 |
+
"rank": 512,
|
| 546 |
+
"sparse_nnz": 0
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"module_path": "model.layers.11.self_attn.k_proj",
|
| 550 |
+
"in_features": 4096,
|
| 551 |
+
"out_features": 4096,
|
| 552 |
+
"rank": 512,
|
| 553 |
+
"sparse_nnz": 0
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"module_path": "model.layers.11.self_attn.v_proj",
|
| 557 |
+
"in_features": 4096,
|
| 558 |
+
"out_features": 4096,
|
| 559 |
+
"rank": 512,
|
| 560 |
+
"sparse_nnz": 0
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"module_path": "model.layers.11.self_attn.o_proj",
|
| 564 |
+
"in_features": 4096,
|
| 565 |
+
"out_features": 4096,
|
| 566 |
+
"rank": 512,
|
| 567 |
+
"sparse_nnz": 0
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"module_path": "model.layers.11.mlp.gate_proj",
|
| 571 |
+
"in_features": 4096,
|
| 572 |
+
"out_features": 11008,
|
| 573 |
+
"rank": 2132,
|
| 574 |
+
"sparse_nnz": 0
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"module_path": "model.layers.11.mlp.up_proj",
|
| 578 |
+
"in_features": 4096,
|
| 579 |
+
"out_features": 11008,
|
| 580 |
+
"rank": 2132,
|
| 581 |
+
"sparse_nnz": 0
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"module_path": "model.layers.11.mlp.down_proj",
|
| 585 |
+
"in_features": 11008,
|
| 586 |
+
"out_features": 4096,
|
| 587 |
+
"rank": 2132,
|
| 588 |
+
"sparse_nnz": 0
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"module_path": "model.layers.12.self_attn.q_proj",
|
| 592 |
+
"in_features": 4096,
|
| 593 |
+
"out_features": 4096,
|
| 594 |
+
"rank": 512,
|
| 595 |
+
"sparse_nnz": 0
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"module_path": "model.layers.12.self_attn.k_proj",
|
| 599 |
+
"in_features": 4096,
|
| 600 |
+
"out_features": 4096,
|
| 601 |
+
"rank": 512,
|
| 602 |
+
"sparse_nnz": 0
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"module_path": "model.layers.12.self_attn.v_proj",
|
| 606 |
+
"in_features": 4096,
|
| 607 |
+
"out_features": 4096,
|
| 608 |
+
"rank": 512,
|
| 609 |
+
"sparse_nnz": 0
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"module_path": "model.layers.12.self_attn.o_proj",
|
| 613 |
+
"in_features": 4096,
|
| 614 |
+
"out_features": 4096,
|
| 615 |
+
"rank": 512,
|
| 616 |
+
"sparse_nnz": 0
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"module_path": "model.layers.12.mlp.gate_proj",
|
| 620 |
+
"in_features": 4096,
|
| 621 |
+
"out_features": 11008,
|
| 622 |
+
"rank": 2132,
|
| 623 |
+
"sparse_nnz": 0
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"module_path": "model.layers.12.mlp.up_proj",
|
| 627 |
+
"in_features": 4096,
|
| 628 |
+
"out_features": 11008,
|
| 629 |
+
"rank": 2132,
|
| 630 |
+
"sparse_nnz": 0
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"module_path": "model.layers.12.mlp.down_proj",
|
| 634 |
+
"in_features": 11008,
|
| 635 |
+
"out_features": 4096,
|
| 636 |
+
"rank": 2132,
|
| 637 |
+
"sparse_nnz": 0
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"module_path": "model.layers.13.self_attn.q_proj",
|
| 641 |
+
"in_features": 4096,
|
| 642 |
+
"out_features": 4096,
|
| 643 |
+
"rank": 512,
|
| 644 |
+
"sparse_nnz": 0
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"module_path": "model.layers.13.self_attn.k_proj",
|
| 648 |
+
"in_features": 4096,
|
| 649 |
+
"out_features": 4096,
|
| 650 |
+
"rank": 512,
|
| 651 |
+
"sparse_nnz": 0
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"module_path": "model.layers.13.self_attn.v_proj",
|
| 655 |
+
"in_features": 4096,
|
| 656 |
+
"out_features": 4096,
|
| 657 |
+
"rank": 512,
|
| 658 |
+
"sparse_nnz": 0
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"module_path": "model.layers.13.self_attn.o_proj",
|
| 662 |
+
"in_features": 4096,
|
| 663 |
+
"out_features": 4096,
|
| 664 |
+
"rank": 512,
|
| 665 |
+
"sparse_nnz": 0
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"module_path": "model.layers.13.mlp.gate_proj",
|
| 669 |
+
"in_features": 4096,
|
| 670 |
+
"out_features": 11008,
|
| 671 |
+
"rank": 2132,
|
| 672 |
+
"sparse_nnz": 0
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"module_path": "model.layers.13.mlp.up_proj",
|
| 676 |
+
"in_features": 4096,
|
| 677 |
+
"out_features": 11008,
|
| 678 |
+
"rank": 2132,
|
| 679 |
+
"sparse_nnz": 0
|
| 680 |
+
},
|
| 681 |
+
{
|
| 682 |
+
"module_path": "model.layers.13.mlp.down_proj",
|
| 683 |
+
"in_features": 11008,
|
| 684 |
+
"out_features": 4096,
|
| 685 |
+
"rank": 2132,
|
| 686 |
+
"sparse_nnz": 0
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"module_path": "model.layers.14.self_attn.q_proj",
|
| 690 |
+
"in_features": 4096,
|
| 691 |
+
"out_features": 4096,
|
| 692 |
+
"rank": 512,
|
| 693 |
+
"sparse_nnz": 0
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"module_path": "model.layers.14.self_attn.k_proj",
|
| 697 |
+
"in_features": 4096,
|
| 698 |
+
"out_features": 4096,
|
| 699 |
+
"rank": 512,
|
| 700 |
+
"sparse_nnz": 0
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"module_path": "model.layers.14.self_attn.v_proj",
|
| 704 |
+
"in_features": 4096,
|
| 705 |
+
"out_features": 4096,
|
| 706 |
+
"rank": 512,
|
| 707 |
+
"sparse_nnz": 0
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"module_path": "model.layers.14.self_attn.o_proj",
|
| 711 |
+
"in_features": 4096,
|
| 712 |
+
"out_features": 4096,
|
| 713 |
+
"rank": 512,
|
| 714 |
+
"sparse_nnz": 0
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"module_path": "model.layers.14.mlp.gate_proj",
|
| 718 |
+
"in_features": 4096,
|
| 719 |
+
"out_features": 11008,
|
| 720 |
+
"rank": 2132,
|
| 721 |
+
"sparse_nnz": 0
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"module_path": "model.layers.14.mlp.up_proj",
|
| 725 |
+
"in_features": 4096,
|
| 726 |
+
"out_features": 11008,
|
| 727 |
+
"rank": 2132,
|
| 728 |
+
"sparse_nnz": 0
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"module_path": "model.layers.14.mlp.down_proj",
|
| 732 |
+
"in_features": 11008,
|
| 733 |
+
"out_features": 4096,
|
| 734 |
+
"rank": 2132,
|
| 735 |
+
"sparse_nnz": 0
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"module_path": "model.layers.15.self_attn.q_proj",
|
| 739 |
+
"in_features": 4096,
|
| 740 |
+
"out_features": 4096,
|
| 741 |
+
"rank": 512,
|
| 742 |
+
"sparse_nnz": 0
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"module_path": "model.layers.15.self_attn.k_proj",
|
| 746 |
+
"in_features": 4096,
|
| 747 |
+
"out_features": 4096,
|
| 748 |
+
"rank": 512,
|
| 749 |
+
"sparse_nnz": 0
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"module_path": "model.layers.15.self_attn.v_proj",
|
| 753 |
+
"in_features": 4096,
|
| 754 |
+
"out_features": 4096,
|
| 755 |
+
"rank": 512,
|
| 756 |
+
"sparse_nnz": 0
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"module_path": "model.layers.15.self_attn.o_proj",
|
| 760 |
+
"in_features": 4096,
|
| 761 |
+
"out_features": 4096,
|
| 762 |
+
"rank": 512,
|
| 763 |
+
"sparse_nnz": 0
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"module_path": "model.layers.15.mlp.gate_proj",
|
| 767 |
+
"in_features": 4096,
|
| 768 |
+
"out_features": 11008,
|
| 769 |
+
"rank": 2132,
|
| 770 |
+
"sparse_nnz": 0
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"module_path": "model.layers.15.mlp.up_proj",
|
| 774 |
+
"in_features": 4096,
|
| 775 |
+
"out_features": 11008,
|
| 776 |
+
"rank": 2132,
|
| 777 |
+
"sparse_nnz": 0
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"module_path": "model.layers.15.mlp.down_proj",
|
| 781 |
+
"in_features": 11008,
|
| 782 |
+
"out_features": 4096,
|
| 783 |
+
"rank": 2132,
|
| 784 |
+
"sparse_nnz": 0
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"module_path": "model.layers.16.self_attn.q_proj",
|
| 788 |
+
"in_features": 4096,
|
| 789 |
+
"out_features": 4096,
|
| 790 |
+
"rank": 512,
|
| 791 |
+
"sparse_nnz": 0
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"module_path": "model.layers.16.self_attn.k_proj",
|
| 795 |
+
"in_features": 4096,
|
| 796 |
+
"out_features": 4096,
|
| 797 |
+
"rank": 512,
|
| 798 |
+
"sparse_nnz": 0
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"module_path": "model.layers.16.self_attn.v_proj",
|
| 802 |
+
"in_features": 4096,
|
| 803 |
+
"out_features": 4096,
|
| 804 |
+
"rank": 512,
|
| 805 |
+
"sparse_nnz": 0
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"module_path": "model.layers.16.self_attn.o_proj",
|
| 809 |
+
"in_features": 4096,
|
| 810 |
+
"out_features": 4096,
|
| 811 |
+
"rank": 512,
|
| 812 |
+
"sparse_nnz": 0
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"module_path": "model.layers.16.mlp.gate_proj",
|
| 816 |
+
"in_features": 4096,
|
| 817 |
+
"out_features": 11008,
|
| 818 |
+
"rank": 2132,
|
| 819 |
+
"sparse_nnz": 0
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"module_path": "model.layers.16.mlp.up_proj",
|
| 823 |
+
"in_features": 4096,
|
| 824 |
+
"out_features": 11008,
|
| 825 |
+
"rank": 2132,
|
| 826 |
+
"sparse_nnz": 0
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"module_path": "model.layers.16.mlp.down_proj",
|
| 830 |
+
"in_features": 11008,
|
| 831 |
+
"out_features": 4096,
|
| 832 |
+
"rank": 2132,
|
| 833 |
+
"sparse_nnz": 0
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"module_path": "model.layers.17.self_attn.q_proj",
|
| 837 |
+
"in_features": 4096,
|
| 838 |
+
"out_features": 4096,
|
| 839 |
+
"rank": 512,
|
| 840 |
+
"sparse_nnz": 0
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"module_path": "model.layers.17.self_attn.k_proj",
|
| 844 |
+
"in_features": 4096,
|
| 845 |
+
"out_features": 4096,
|
| 846 |
+
"rank": 512,
|
| 847 |
+
"sparse_nnz": 0
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"module_path": "model.layers.17.self_attn.v_proj",
|
| 851 |
+
"in_features": 4096,
|
| 852 |
+
"out_features": 4096,
|
| 853 |
+
"rank": 512,
|
| 854 |
+
"sparse_nnz": 0
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"module_path": "model.layers.17.self_attn.o_proj",
|
| 858 |
+
"in_features": 4096,
|
| 859 |
+
"out_features": 4096,
|
| 860 |
+
"rank": 512,
|
| 861 |
+
"sparse_nnz": 0
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"module_path": "model.layers.17.mlp.gate_proj",
|
| 865 |
+
"in_features": 4096,
|
| 866 |
+
"out_features": 11008,
|
| 867 |
+
"rank": 2132,
|
| 868 |
+
"sparse_nnz": 0
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"module_path": "model.layers.17.mlp.up_proj",
|
| 872 |
+
"in_features": 4096,
|
| 873 |
+
"out_features": 11008,
|
| 874 |
+
"rank": 2132,
|
| 875 |
+
"sparse_nnz": 0
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"module_path": "model.layers.17.mlp.down_proj",
|
| 879 |
+
"in_features": 11008,
|
| 880 |
+
"out_features": 4096,
|
| 881 |
+
"rank": 2132,
|
| 882 |
+
"sparse_nnz": 0
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"module_path": "model.layers.18.self_attn.q_proj",
|
| 886 |
+
"in_features": 4096,
|
| 887 |
+
"out_features": 4096,
|
| 888 |
+
"rank": 512,
|
| 889 |
+
"sparse_nnz": 0
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"module_path": "model.layers.18.self_attn.k_proj",
|
| 893 |
+
"in_features": 4096,
|
| 894 |
+
"out_features": 4096,
|
| 895 |
+
"rank": 512,
|
| 896 |
+
"sparse_nnz": 0
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"module_path": "model.layers.18.self_attn.v_proj",
|
| 900 |
+
"in_features": 4096,
|
| 901 |
+
"out_features": 4096,
|
| 902 |
+
"rank": 512,
|
| 903 |
+
"sparse_nnz": 0
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"module_path": "model.layers.18.self_attn.o_proj",
|
| 907 |
+
"in_features": 4096,
|
| 908 |
+
"out_features": 4096,
|
| 909 |
+
"rank": 512,
|
| 910 |
+
"sparse_nnz": 0
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"module_path": "model.layers.18.mlp.gate_proj",
|
| 914 |
+
"in_features": 4096,
|
| 915 |
+
"out_features": 11008,
|
| 916 |
+
"rank": 2132,
|
| 917 |
+
"sparse_nnz": 0
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"module_path": "model.layers.18.mlp.up_proj",
|
| 921 |
+
"in_features": 4096,
|
| 922 |
+
"out_features": 11008,
|
| 923 |
+
"rank": 2132,
|
| 924 |
+
"sparse_nnz": 0
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"module_path": "model.layers.18.mlp.down_proj",
|
| 928 |
+
"in_features": 11008,
|
| 929 |
+
"out_features": 4096,
|
| 930 |
+
"rank": 2132,
|
| 931 |
+
"sparse_nnz": 0
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"module_path": "model.layers.19.self_attn.q_proj",
|
| 935 |
+
"in_features": 4096,
|
| 936 |
+
"out_features": 4096,
|
| 937 |
+
"rank": 512,
|
| 938 |
+
"sparse_nnz": 0
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"module_path": "model.layers.19.self_attn.k_proj",
|
| 942 |
+
"in_features": 4096,
|
| 943 |
+
"out_features": 4096,
|
| 944 |
+
"rank": 512,
|
| 945 |
+
"sparse_nnz": 0
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"module_path": "model.layers.19.self_attn.v_proj",
|
| 949 |
+
"in_features": 4096,
|
| 950 |
+
"out_features": 4096,
|
| 951 |
+
"rank": 512,
|
| 952 |
+
"sparse_nnz": 0
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"module_path": "model.layers.19.self_attn.o_proj",
|
| 956 |
+
"in_features": 4096,
|
| 957 |
+
"out_features": 4096,
|
| 958 |
+
"rank": 512,
|
| 959 |
+
"sparse_nnz": 0
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"module_path": "model.layers.19.mlp.gate_proj",
|
| 963 |
+
"in_features": 4096,
|
| 964 |
+
"out_features": 11008,
|
| 965 |
+
"rank": 2132,
|
| 966 |
+
"sparse_nnz": 0
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"module_path": "model.layers.19.mlp.up_proj",
|
| 970 |
+
"in_features": 4096,
|
| 971 |
+
"out_features": 11008,
|
| 972 |
+
"rank": 2132,
|
| 973 |
+
"sparse_nnz": 0
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"module_path": "model.layers.19.mlp.down_proj",
|
| 977 |
+
"in_features": 11008,
|
| 978 |
+
"out_features": 4096,
|
| 979 |
+
"rank": 2132,
|
| 980 |
+
"sparse_nnz": 0
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"module_path": "model.layers.20.self_attn.q_proj",
|
| 984 |
+
"in_features": 4096,
|
| 985 |
+
"out_features": 4096,
|
| 986 |
+
"rank": 512,
|
| 987 |
+
"sparse_nnz": 0
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"module_path": "model.layers.20.self_attn.k_proj",
|
| 991 |
+
"in_features": 4096,
|
| 992 |
+
"out_features": 4096,
|
| 993 |
+
"rank": 512,
|
| 994 |
+
"sparse_nnz": 0
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"module_path": "model.layers.20.self_attn.v_proj",
|
| 998 |
+
"in_features": 4096,
|
| 999 |
+
"out_features": 4096,
|
| 1000 |
+
"rank": 512,
|
| 1001 |
+
"sparse_nnz": 0
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"module_path": "model.layers.20.self_attn.o_proj",
|
| 1005 |
+
"in_features": 4096,
|
| 1006 |
+
"out_features": 4096,
|
| 1007 |
+
"rank": 512,
|
| 1008 |
+
"sparse_nnz": 0
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"module_path": "model.layers.20.mlp.gate_proj",
|
| 1012 |
+
"in_features": 4096,
|
| 1013 |
+
"out_features": 11008,
|
| 1014 |
+
"rank": 2132,
|
| 1015 |
+
"sparse_nnz": 0
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"module_path": "model.layers.20.mlp.up_proj",
|
| 1019 |
+
"in_features": 4096,
|
| 1020 |
+
"out_features": 11008,
|
| 1021 |
+
"rank": 2132,
|
| 1022 |
+
"sparse_nnz": 0
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"module_path": "model.layers.20.mlp.down_proj",
|
| 1026 |
+
"in_features": 11008,
|
| 1027 |
+
"out_features": 4096,
|
| 1028 |
+
"rank": 2132,
|
| 1029 |
+
"sparse_nnz": 0
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"module_path": "model.layers.21.self_attn.q_proj",
|
| 1033 |
+
"in_features": 4096,
|
| 1034 |
+
"out_features": 4096,
|
| 1035 |
+
"rank": 512,
|
| 1036 |
+
"sparse_nnz": 0
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"module_path": "model.layers.21.self_attn.k_proj",
|
| 1040 |
+
"in_features": 4096,
|
| 1041 |
+
"out_features": 4096,
|
| 1042 |
+
"rank": 512,
|
| 1043 |
+
"sparse_nnz": 0
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"module_path": "model.layers.21.self_attn.v_proj",
|
| 1047 |
+
"in_features": 4096,
|
| 1048 |
+
"out_features": 4096,
|
| 1049 |
+
"rank": 512,
|
| 1050 |
+
"sparse_nnz": 0
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"module_path": "model.layers.21.self_attn.o_proj",
|
| 1054 |
+
"in_features": 4096,
|
| 1055 |
+
"out_features": 4096,
|
| 1056 |
+
"rank": 512,
|
| 1057 |
+
"sparse_nnz": 0
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"module_path": "model.layers.21.mlp.gate_proj",
|
| 1061 |
+
"in_features": 4096,
|
| 1062 |
+
"out_features": 11008,
|
| 1063 |
+
"rank": 2132,
|
| 1064 |
+
"sparse_nnz": 0
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"module_path": "model.layers.21.mlp.up_proj",
|
| 1068 |
+
"in_features": 4096,
|
| 1069 |
+
"out_features": 11008,
|
| 1070 |
+
"rank": 2132,
|
| 1071 |
+
"sparse_nnz": 0
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"module_path": "model.layers.21.mlp.down_proj",
|
| 1075 |
+
"in_features": 11008,
|
| 1076 |
+
"out_features": 4096,
|
| 1077 |
+
"rank": 2132,
|
| 1078 |
+
"sparse_nnz": 0
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"module_path": "model.layers.22.self_attn.q_proj",
|
| 1082 |
+
"in_features": 4096,
|
| 1083 |
+
"out_features": 4096,
|
| 1084 |
+
"rank": 512,
|
| 1085 |
+
"sparse_nnz": 0
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"module_path": "model.layers.22.self_attn.k_proj",
|
| 1089 |
+
"in_features": 4096,
|
| 1090 |
+
"out_features": 4096,
|
| 1091 |
+
"rank": 512,
|
| 1092 |
+
"sparse_nnz": 0
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"module_path": "model.layers.22.self_attn.v_proj",
|
| 1096 |
+
"in_features": 4096,
|
| 1097 |
+
"out_features": 4096,
|
| 1098 |
+
"rank": 512,
|
| 1099 |
+
"sparse_nnz": 0
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"module_path": "model.layers.22.self_attn.o_proj",
|
| 1103 |
+
"in_features": 4096,
|
| 1104 |
+
"out_features": 4096,
|
| 1105 |
+
"rank": 512,
|
| 1106 |
+
"sparse_nnz": 0
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"module_path": "model.layers.22.mlp.gate_proj",
|
| 1110 |
+
"in_features": 4096,
|
| 1111 |
+
"out_features": 11008,
|
| 1112 |
+
"rank": 2132,
|
| 1113 |
+
"sparse_nnz": 0
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"module_path": "model.layers.22.mlp.up_proj",
|
| 1117 |
+
"in_features": 4096,
|
| 1118 |
+
"out_features": 11008,
|
| 1119 |
+
"rank": 2132,
|
| 1120 |
+
"sparse_nnz": 0
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"module_path": "model.layers.22.mlp.down_proj",
|
| 1124 |
+
"in_features": 11008,
|
| 1125 |
+
"out_features": 4096,
|
| 1126 |
+
"rank": 2132,
|
| 1127 |
+
"sparse_nnz": 0
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"module_path": "model.layers.23.self_attn.q_proj",
|
| 1131 |
+
"in_features": 4096,
|
| 1132 |
+
"out_features": 4096,
|
| 1133 |
+
"rank": 512,
|
| 1134 |
+
"sparse_nnz": 0
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"module_path": "model.layers.23.self_attn.k_proj",
|
| 1138 |
+
"in_features": 4096,
|
| 1139 |
+
"out_features": 4096,
|
| 1140 |
+
"rank": 512,
|
| 1141 |
+
"sparse_nnz": 0
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"module_path": "model.layers.23.self_attn.v_proj",
|
| 1145 |
+
"in_features": 4096,
|
| 1146 |
+
"out_features": 4096,
|
| 1147 |
+
"rank": 512,
|
| 1148 |
+
"sparse_nnz": 0
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"module_path": "model.layers.23.self_attn.o_proj",
|
| 1152 |
+
"in_features": 4096,
|
| 1153 |
+
"out_features": 4096,
|
| 1154 |
+
"rank": 512,
|
| 1155 |
+
"sparse_nnz": 0
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"module_path": "model.layers.23.mlp.gate_proj",
|
| 1159 |
+
"in_features": 4096,
|
| 1160 |
+
"out_features": 11008,
|
| 1161 |
+
"rank": 2132,
|
| 1162 |
+
"sparse_nnz": 0
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"module_path": "model.layers.23.mlp.up_proj",
|
| 1166 |
+
"in_features": 4096,
|
| 1167 |
+
"out_features": 11008,
|
| 1168 |
+
"rank": 2132,
|
| 1169 |
+
"sparse_nnz": 0
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"module_path": "model.layers.23.mlp.down_proj",
|
| 1173 |
+
"in_features": 11008,
|
| 1174 |
+
"out_features": 4096,
|
| 1175 |
+
"rank": 2132,
|
| 1176 |
+
"sparse_nnz": 0
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"module_path": "model.layers.24.self_attn.q_proj",
|
| 1180 |
+
"in_features": 4096,
|
| 1181 |
+
"out_features": 4096,
|
| 1182 |
+
"rank": 512,
|
| 1183 |
+
"sparse_nnz": 0
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"module_path": "model.layers.24.self_attn.k_proj",
|
| 1187 |
+
"in_features": 4096,
|
| 1188 |
+
"out_features": 4096,
|
| 1189 |
+
"rank": 512,
|
| 1190 |
+
"sparse_nnz": 0
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"module_path": "model.layers.24.self_attn.v_proj",
|
| 1194 |
+
"in_features": 4096,
|
| 1195 |
+
"out_features": 4096,
|
| 1196 |
+
"rank": 512,
|
| 1197 |
+
"sparse_nnz": 0
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"module_path": "model.layers.24.self_attn.o_proj",
|
| 1201 |
+
"in_features": 4096,
|
| 1202 |
+
"out_features": 4096,
|
| 1203 |
+
"rank": 512,
|
| 1204 |
+
"sparse_nnz": 0
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"module_path": "model.layers.24.mlp.gate_proj",
|
| 1208 |
+
"in_features": 4096,
|
| 1209 |
+
"out_features": 11008,
|
| 1210 |
+
"rank": 2132,
|
| 1211 |
+
"sparse_nnz": 0
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"module_path": "model.layers.24.mlp.up_proj",
|
| 1215 |
+
"in_features": 4096,
|
| 1216 |
+
"out_features": 11008,
|
| 1217 |
+
"rank": 2132,
|
| 1218 |
+
"sparse_nnz": 0
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"module_path": "model.layers.24.mlp.down_proj",
|
| 1222 |
+
"in_features": 11008,
|
| 1223 |
+
"out_features": 4096,
|
| 1224 |
+
"rank": 2132,
|
| 1225 |
+
"sparse_nnz": 0
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"module_path": "model.layers.25.self_attn.q_proj",
|
| 1229 |
+
"in_features": 4096,
|
| 1230 |
+
"out_features": 4096,
|
| 1231 |
+
"rank": 512,
|
| 1232 |
+
"sparse_nnz": 0
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"module_path": "model.layers.25.self_attn.k_proj",
|
| 1236 |
+
"in_features": 4096,
|
| 1237 |
+
"out_features": 4096,
|
| 1238 |
+
"rank": 512,
|
| 1239 |
+
"sparse_nnz": 0
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"module_path": "model.layers.25.self_attn.v_proj",
|
| 1243 |
+
"in_features": 4096,
|
| 1244 |
+
"out_features": 4096,
|
| 1245 |
+
"rank": 512,
|
| 1246 |
+
"sparse_nnz": 0
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"module_path": "model.layers.25.self_attn.o_proj",
|
| 1250 |
+
"in_features": 4096,
|
| 1251 |
+
"out_features": 4096,
|
| 1252 |
+
"rank": 512,
|
| 1253 |
+
"sparse_nnz": 0
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"module_path": "model.layers.25.mlp.gate_proj",
|
| 1257 |
+
"in_features": 4096,
|
| 1258 |
+
"out_features": 11008,
|
| 1259 |
+
"rank": 2132,
|
| 1260 |
+
"sparse_nnz": 0
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"module_path": "model.layers.25.mlp.up_proj",
|
| 1264 |
+
"in_features": 4096,
|
| 1265 |
+
"out_features": 11008,
|
| 1266 |
+
"rank": 2132,
|
| 1267 |
+
"sparse_nnz": 0
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"module_path": "model.layers.25.mlp.down_proj",
|
| 1271 |
+
"in_features": 11008,
|
| 1272 |
+
"out_features": 4096,
|
| 1273 |
+
"rank": 2132,
|
| 1274 |
+
"sparse_nnz": 0
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"module_path": "model.layers.26.self_attn.q_proj",
|
| 1278 |
+
"in_features": 4096,
|
| 1279 |
+
"out_features": 4096,
|
| 1280 |
+
"rank": 512,
|
| 1281 |
+
"sparse_nnz": 0
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"module_path": "model.layers.26.self_attn.k_proj",
|
| 1285 |
+
"in_features": 4096,
|
| 1286 |
+
"out_features": 4096,
|
| 1287 |
+
"rank": 512,
|
| 1288 |
+
"sparse_nnz": 0
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"module_path": "model.layers.26.self_attn.v_proj",
|
| 1292 |
+
"in_features": 4096,
|
| 1293 |
+
"out_features": 4096,
|
| 1294 |
+
"rank": 512,
|
| 1295 |
+
"sparse_nnz": 0
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"module_path": "model.layers.26.self_attn.o_proj",
|
| 1299 |
+
"in_features": 4096,
|
| 1300 |
+
"out_features": 4096,
|
| 1301 |
+
"rank": 512,
|
| 1302 |
+
"sparse_nnz": 0
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"module_path": "model.layers.26.mlp.gate_proj",
|
| 1306 |
+
"in_features": 4096,
|
| 1307 |
+
"out_features": 11008,
|
| 1308 |
+
"rank": 2132,
|
| 1309 |
+
"sparse_nnz": 0
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"module_path": "model.layers.26.mlp.up_proj",
|
| 1313 |
+
"in_features": 4096,
|
| 1314 |
+
"out_features": 11008,
|
| 1315 |
+
"rank": 2132,
|
| 1316 |
+
"sparse_nnz": 0
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"module_path": "model.layers.26.mlp.down_proj",
|
| 1320 |
+
"in_features": 11008,
|
| 1321 |
+
"out_features": 4096,
|
| 1322 |
+
"rank": 2132,
|
| 1323 |
+
"sparse_nnz": 0
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"module_path": "model.layers.27.self_attn.q_proj",
|
| 1327 |
+
"in_features": 4096,
|
| 1328 |
+
"out_features": 4096,
|
| 1329 |
+
"rank": 512,
|
| 1330 |
+
"sparse_nnz": 0
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"module_path": "model.layers.27.self_attn.k_proj",
|
| 1334 |
+
"in_features": 4096,
|
| 1335 |
+
"out_features": 4096,
|
| 1336 |
+
"rank": 512,
|
| 1337 |
+
"sparse_nnz": 0
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"module_path": "model.layers.27.self_attn.v_proj",
|
| 1341 |
+
"in_features": 4096,
|
| 1342 |
+
"out_features": 4096,
|
| 1343 |
+
"rank": 512,
|
| 1344 |
+
"sparse_nnz": 0
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"module_path": "model.layers.27.self_attn.o_proj",
|
| 1348 |
+
"in_features": 4096,
|
| 1349 |
+
"out_features": 4096,
|
| 1350 |
+
"rank": 512,
|
| 1351 |
+
"sparse_nnz": 0
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"module_path": "model.layers.27.mlp.gate_proj",
|
| 1355 |
+
"in_features": 4096,
|
| 1356 |
+
"out_features": 11008,
|
| 1357 |
+
"rank": 2132,
|
| 1358 |
+
"sparse_nnz": 0
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"module_path": "model.layers.27.mlp.up_proj",
|
| 1362 |
+
"in_features": 4096,
|
| 1363 |
+
"out_features": 11008,
|
| 1364 |
+
"rank": 2132,
|
| 1365 |
+
"sparse_nnz": 0
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"module_path": "model.layers.27.mlp.down_proj",
|
| 1369 |
+
"in_features": 11008,
|
| 1370 |
+
"out_features": 4096,
|
| 1371 |
+
"rank": 2132,
|
| 1372 |
+
"sparse_nnz": 0
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"module_path": "model.layers.28.self_attn.q_proj",
|
| 1376 |
+
"in_features": 4096,
|
| 1377 |
+
"out_features": 4096,
|
| 1378 |
+
"rank": 512,
|
| 1379 |
+
"sparse_nnz": 0
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"module_path": "model.layers.28.self_attn.k_proj",
|
| 1383 |
+
"in_features": 4096,
|
| 1384 |
+
"out_features": 4096,
|
| 1385 |
+
"rank": 512,
|
| 1386 |
+
"sparse_nnz": 0
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"module_path": "model.layers.28.self_attn.v_proj",
|
| 1390 |
+
"in_features": 4096,
|
| 1391 |
+
"out_features": 4096,
|
| 1392 |
+
"rank": 512,
|
| 1393 |
+
"sparse_nnz": 0
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"module_path": "model.layers.28.self_attn.o_proj",
|
| 1397 |
+
"in_features": 4096,
|
| 1398 |
+
"out_features": 4096,
|
| 1399 |
+
"rank": 512,
|
| 1400 |
+
"sparse_nnz": 0
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"module_path": "model.layers.28.mlp.gate_proj",
|
| 1404 |
+
"in_features": 4096,
|
| 1405 |
+
"out_features": 11008,
|
| 1406 |
+
"rank": 2132,
|
| 1407 |
+
"sparse_nnz": 0
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"module_path": "model.layers.28.mlp.up_proj",
|
| 1411 |
+
"in_features": 4096,
|
| 1412 |
+
"out_features": 11008,
|
| 1413 |
+
"rank": 2132,
|
| 1414 |
+
"sparse_nnz": 0
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"module_path": "model.layers.28.mlp.down_proj",
|
| 1418 |
+
"in_features": 11008,
|
| 1419 |
+
"out_features": 4096,
|
| 1420 |
+
"rank": 2132,
|
| 1421 |
+
"sparse_nnz": 0
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"module_path": "model.layers.29.self_attn.q_proj",
|
| 1425 |
+
"in_features": 4096,
|
| 1426 |
+
"out_features": 4096,
|
| 1427 |
+
"rank": 512,
|
| 1428 |
+
"sparse_nnz": 0
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"module_path": "model.layers.29.self_attn.k_proj",
|
| 1432 |
+
"in_features": 4096,
|
| 1433 |
+
"out_features": 4096,
|
| 1434 |
+
"rank": 512,
|
| 1435 |
+
"sparse_nnz": 0
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"module_path": "model.layers.29.self_attn.v_proj",
|
| 1439 |
+
"in_features": 4096,
|
| 1440 |
+
"out_features": 4096,
|
| 1441 |
+
"rank": 512,
|
| 1442 |
+
"sparse_nnz": 0
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"module_path": "model.layers.29.self_attn.o_proj",
|
| 1446 |
+
"in_features": 4096,
|
| 1447 |
+
"out_features": 4096,
|
| 1448 |
+
"rank": 512,
|
| 1449 |
+
"sparse_nnz": 0
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"module_path": "model.layers.29.mlp.gate_proj",
|
| 1453 |
+
"in_features": 4096,
|
| 1454 |
+
"out_features": 11008,
|
| 1455 |
+
"rank": 2132,
|
| 1456 |
+
"sparse_nnz": 0
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"module_path": "model.layers.29.mlp.up_proj",
|
| 1460 |
+
"in_features": 4096,
|
| 1461 |
+
"out_features": 11008,
|
| 1462 |
+
"rank": 2132,
|
| 1463 |
+
"sparse_nnz": 0
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"module_path": "model.layers.29.mlp.down_proj",
|
| 1467 |
+
"in_features": 11008,
|
| 1468 |
+
"out_features": 4096,
|
| 1469 |
+
"rank": 2132,
|
| 1470 |
+
"sparse_nnz": 0
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"module_path": "model.layers.30.self_attn.q_proj",
|
| 1474 |
+
"in_features": 4096,
|
| 1475 |
+
"out_features": 4096,
|
| 1476 |
+
"rank": 512,
|
| 1477 |
+
"sparse_nnz": 0
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"module_path": "model.layers.30.self_attn.k_proj",
|
| 1481 |
+
"in_features": 4096,
|
| 1482 |
+
"out_features": 4096,
|
| 1483 |
+
"rank": 512,
|
| 1484 |
+
"sparse_nnz": 0
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"module_path": "model.layers.30.self_attn.v_proj",
|
| 1488 |
+
"in_features": 4096,
|
| 1489 |
+
"out_features": 4096,
|
| 1490 |
+
"rank": 512,
|
| 1491 |
+
"sparse_nnz": 0
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"module_path": "model.layers.30.self_attn.o_proj",
|
| 1495 |
+
"in_features": 4096,
|
| 1496 |
+
"out_features": 4096,
|
| 1497 |
+
"rank": 512,
|
| 1498 |
+
"sparse_nnz": 0
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"module_path": "model.layers.30.mlp.gate_proj",
|
| 1502 |
+
"in_features": 4096,
|
| 1503 |
+
"out_features": 11008,
|
| 1504 |
+
"rank": 2132,
|
| 1505 |
+
"sparse_nnz": 0
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"module_path": "model.layers.30.mlp.up_proj",
|
| 1509 |
+
"in_features": 4096,
|
| 1510 |
+
"out_features": 11008,
|
| 1511 |
+
"rank": 2132,
|
| 1512 |
+
"sparse_nnz": 0
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"module_path": "model.layers.30.mlp.down_proj",
|
| 1516 |
+
"in_features": 11008,
|
| 1517 |
+
"out_features": 4096,
|
| 1518 |
+
"rank": 2132,
|
| 1519 |
+
"sparse_nnz": 0
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"module_path": "model.layers.31.self_attn.q_proj",
|
| 1523 |
+
"in_features": 4096,
|
| 1524 |
+
"out_features": 4096,
|
| 1525 |
+
"rank": 512,
|
| 1526 |
+
"sparse_nnz": 0
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"module_path": "model.layers.31.self_attn.k_proj",
|
| 1530 |
+
"in_features": 4096,
|
| 1531 |
+
"out_features": 4096,
|
| 1532 |
+
"rank": 512,
|
| 1533 |
+
"sparse_nnz": 0
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"module_path": "model.layers.31.self_attn.v_proj",
|
| 1537 |
+
"in_features": 4096,
|
| 1538 |
+
"out_features": 4096,
|
| 1539 |
+
"rank": 512,
|
| 1540 |
+
"sparse_nnz": 0
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"module_path": "model.layers.31.self_attn.o_proj",
|
| 1544 |
+
"in_features": 4096,
|
| 1545 |
+
"out_features": 4096,
|
| 1546 |
+
"rank": 512,
|
| 1547 |
+
"sparse_nnz": 0
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"module_path": "model.layers.31.mlp.gate_proj",
|
| 1551 |
+
"in_features": 4096,
|
| 1552 |
+
"out_features": 11008,
|
| 1553 |
+
"rank": 2132,
|
| 1554 |
+
"sparse_nnz": 0
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"module_path": "model.layers.31.mlp.up_proj",
|
| 1558 |
+
"in_features": 4096,
|
| 1559 |
+
"out_features": 11008,
|
| 1560 |
+
"rank": 2132,
|
| 1561 |
+
"sparse_nnz": 0
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"module_path": "model.layers.31.mlp.down_proj",
|
| 1565 |
+
"in_features": 11008,
|
| 1566 |
+
"out_features": 4096,
|
| 1567 |
+
"rank": 2132,
|
| 1568 |
+
"sparse_nnz": 0
|
| 1569 |
+
}
|
| 1570 |
+
]
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bdb242e1e2a049ef05185f2ce352397b1c474a4280611c93f192c30ba69d3504
|
| 3 |
+
size 8900554064
|
modeling_pmre_olmo.py
ADDED
|
@@ -0,0 +1,241 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
"""HuggingFace-compatible model class for PMRE factorized OLMo.
|
| 2 |
+
|
| 3 |
+
This module defines a custom model class that integrates with HuggingFace's
|
| 4 |
+
AutoModelForCausalLM via auto_map in config.json. Users load with:
|
| 5 |
+
|
| 6 |
+
model = AutoModelForCausalLM.from_pretrained(
|
| 7 |
+
"asystemoffields/OLMo-3-3.5B-Think-PMRE",
|
| 8 |
+
trust_remote_code=True,
|
| 9 |
+
torch_dtype=torch.bfloat16,
|
| 10 |
+
)
|
| 11 |
+
|
| 12 |
+
The factorized weights are stored in safetensors format. Each compressed Linear
|
| 13 |
+
becomes a FactorizedLinear with down/up factor matrices, preserving the 2x
|
| 14 |
+
storage reduction and ~2x inference speedup.
|
| 15 |
+
"""
|
| 16 |
+
|
| 17 |
+
from __future__ import annotations
|
| 18 |
+
|
| 19 |
+
import json
|
| 20 |
+
from pathlib import Path
|
| 21 |
+
from typing import Any, Optional, Tuple, Union
|
| 22 |
+
|
| 23 |
+
import torch
|
| 24 |
+
import torch.nn as nn
|
| 25 |
+
from transformers import AutoConfig, AutoModelForCausalLM, PreTrainedModel
|
| 26 |
+
from transformers.modeling_outputs import CausalLMOutputWithPast
|
| 27 |
+
|
| 28 |
+
|
| 29 |
+
class FactorizedLinear(nn.Module):
|
| 30 |
+
"""Low-rank replacement for nn.Linear: x -> up(down(x))."""
|
| 31 |
+
|
| 32 |
+
def __init__(self, in_features: int, out_features: int, rank: int,
|
| 33 |
+
bias: bool = False, device=None, dtype=None):
|
| 34 |
+
super().__init__()
|
| 35 |
+
self.in_features = in_features
|
| 36 |
+
self.out_features = out_features
|
| 37 |
+
self.rank = rank
|
| 38 |
+
self.down = nn.Linear(in_features, rank, bias=False, device=device, dtype=dtype)
|
| 39 |
+
self.up = nn.Linear(rank, out_features, bias=bias, device=device, dtype=dtype)
|
| 40 |
+
|
| 41 |
+
def forward(self, x: torch.Tensor) -> torch.Tensor:
|
| 42 |
+
return self.up(self.down(x))
|
| 43 |
+
|
| 44 |
+
def extra_repr(self) -> str:
|
| 45 |
+
return (f"in_features={self.in_features}, out_features={self.out_features}, "
|
| 46 |
+
f"rank={self.rank}, compression={self.in_features * self.out_features / (self.rank * (self.in_features + self.out_features)):.1f}x")
|
| 47 |
+
|
| 48 |
+
|
| 49 |
+
class FactorizedSparseLinear(nn.Module):
|
| 50 |
+
"""Low-rank linear plus a fixed sparse residual correction."""
|
| 51 |
+
|
| 52 |
+
def __init__(self, in_features: int, out_features: int, rank: int,
|
| 53 |
+
sparse_nnz: int = 0, bias: bool = False, device=None, dtype=None):
|
| 54 |
+
super().__init__()
|
| 55 |
+
self.in_features = in_features
|
| 56 |
+
self.out_features = out_features
|
| 57 |
+
self.rank = rank
|
| 58 |
+
self.down = nn.Linear(in_features, rank, bias=False, device=device, dtype=dtype)
|
| 59 |
+
self.up = nn.Linear(rank, out_features, bias=bias, device=device, dtype=dtype)
|
| 60 |
+
self.register_buffer("sparse_row_idx", torch.zeros(sparse_nnz, dtype=torch.long, device=device))
|
| 61 |
+
self.register_buffer("sparse_col_idx", torch.zeros(sparse_nnz, dtype=torch.long, device=device))
|
| 62 |
+
self.register_buffer("sparse_values", torch.zeros(sparse_nnz, dtype=dtype or torch.bfloat16, device=device))
|
| 63 |
+
self._sparse_cache: Optional[torch.Tensor] = None
|
| 64 |
+
|
| 65 |
+
@property
|
| 66 |
+
def sparse_nnz(self) -> int:
|
| 67 |
+
return int(self.sparse_values.numel())
|
| 68 |
+
|
| 69 |
+
def forward(self, x: torch.Tensor) -> torch.Tensor:
|
| 70 |
+
low_rank = self.up(self.down(x))
|
| 71 |
+
if self.sparse_nnz == 0:
|
| 72 |
+
return low_rank
|
| 73 |
+
|
| 74 |
+
cached = self._sparse_cache
|
| 75 |
+
if cached is None or cached.device != x.device:
|
| 76 |
+
idx = torch.stack([self.sparse_row_idx, self.sparse_col_idx], dim=0)
|
| 77 |
+
cached = torch.sparse_coo_tensor(
|
| 78 |
+
idx, self.sparse_values.float(),
|
| 79 |
+
size=(self.out_features, self.in_features), device=x.device,
|
| 80 |
+
).coalesce()
|
| 81 |
+
self._sparse_cache = cached
|
| 82 |
+
|
| 83 |
+
flat = x.reshape(-1, self.in_features)
|
| 84 |
+
sparse_out = torch.sparse.mm(cached, flat.float().t()).t()
|
| 85 |
+
sparse_out = sparse_out.reshape(*x.shape[:-1], self.out_features).to(dtype=x.dtype)
|
| 86 |
+
return low_rank + sparse_out
|
| 87 |
+
|
| 88 |
+
def extra_repr(self) -> str:
|
| 89 |
+
return (f"in_features={self.in_features}, out_features={self.out_features}, "
|
| 90 |
+
f"rank={self.rank}, sparse_nnz={self.sparse_nnz}")
|
| 91 |
+
|
| 92 |
+
|
| 93 |
+
def _parent_and_attr(model: nn.Module, path: str) -> Tuple[nn.Module, str]:
|
| 94 |
+
parts = path.split(".")
|
| 95 |
+
parent = model.get_submodule(".".join(parts[:-1]))
|
| 96 |
+
return parent, parts[-1]
|
| 97 |
+
|
| 98 |
+
|
| 99 |
+
def apply_factorized_modules(model: nn.Module, metadata: list[dict[str, Any]],
|
| 100 |
+
device=None, dtype=None) -> nn.Module:
|
| 101 |
+
"""Replace target Linear modules with factorized versions based on metadata."""
|
| 102 |
+
for entry in metadata:
|
| 103 |
+
module_path = entry["module_path"]
|
| 104 |
+
rank = int(entry["rank"])
|
| 105 |
+
in_features = int(entry["in_features"])
|
| 106 |
+
out_features = int(entry["out_features"])
|
| 107 |
+
sparse_nnz = int(entry.get("sparse_nnz", 0) or 0)
|
| 108 |
+
|
| 109 |
+
old = model.get_submodule(module_path)
|
| 110 |
+
has_bias = isinstance(old, nn.Linear) and old.bias is not None
|
| 111 |
+
|
| 112 |
+
if sparse_nnz > 0:
|
| 113 |
+
replacement = FactorizedSparseLinear(
|
| 114 |
+
in_features=in_features, out_features=out_features, rank=rank,
|
| 115 |
+
sparse_nnz=sparse_nnz, bias=has_bias, device=device, dtype=dtype,
|
| 116 |
+
)
|
| 117 |
+
else:
|
| 118 |
+
replacement = FactorizedLinear(
|
| 119 |
+
in_features=in_features, out_features=out_features, rank=rank,
|
| 120 |
+
bias=has_bias, device=device, dtype=dtype,
|
| 121 |
+
)
|
| 122 |
+
|
| 123 |
+
parent, attr = _parent_and_attr(model, module_path)
|
| 124 |
+
setattr(parent, attr, replacement)
|
| 125 |
+
return model
|
| 126 |
+
|
| 127 |
+
|
| 128 |
+
class PmreOlmoForCausalLM(PreTrainedModel):
|
| 129 |
+
"""HuggingFace-compatible wrapper for PMRE factorized OLMo models.
|
| 130 |
+
|
| 131 |
+
This class handles:
|
| 132 |
+
1. Building the base OLMo architecture
|
| 133 |
+
2. Replacing compressed modules with FactorizedLinear
|
| 134 |
+
3. Loading the factorized state dict from safetensors/bin files
|
| 135 |
+
4. Standard generate() / forward() interface
|
| 136 |
+
"""
|
| 137 |
+
|
| 138 |
+
config_class = None # Will use the base model's config
|
| 139 |
+
supports_gradient_checkpointing = True
|
| 140 |
+
_no_split_modules = ["OLMo2DecoderLayer"]
|
| 141 |
+
|
| 142 |
+
def __init__(self, config, **kwargs):
|
| 143 |
+
super().__init__(config, **kwargs)
|
| 144 |
+
|
| 145 |
+
# Build the base OLMo model
|
| 146 |
+
base_config = AutoConfig.from_dict(config.to_dict())
|
| 147 |
+
# Remove our custom fields before building base model
|
| 148 |
+
for field in ("pmre_metadata", "auto_map", "model_type_override"):
|
| 149 |
+
if hasattr(base_config, field):
|
| 150 |
+
delattr(base_config, field)
|
| 151 |
+
|
| 152 |
+
self.model_impl = AutoModelForCausalLM.from_config(
|
| 153 |
+
base_config, trust_remote_code=True,
|
| 154 |
+
)
|
| 155 |
+
|
| 156 |
+
# Apply factorized module replacements
|
| 157 |
+
metadata = getattr(config, "pmre_metadata", None)
|
| 158 |
+
if metadata:
|
| 159 |
+
apply_factorized_modules(self.model_impl, metadata, dtype=kwargs.get("torch_dtype"))
|
| 160 |
+
|
| 161 |
+
def forward(self, **kwargs) -> CausalLMOutputWithPast:
|
| 162 |
+
return self.model_impl(**kwargs)
|
| 163 |
+
|
| 164 |
+
def generate(self, *args, **kwargs):
|
| 165 |
+
return self.model_impl.generate(*args, **kwargs)
|
| 166 |
+
|
| 167 |
+
def get_input_embeddings(self):
|
| 168 |
+
return self.model_impl.get_input_embeddings()
|
| 169 |
+
|
| 170 |
+
def set_input_embeddings(self, value):
|
| 171 |
+
self.model_impl.set_input_embeddings(value)
|
| 172 |
+
|
| 173 |
+
def get_output_embeddings(self):
|
| 174 |
+
return self.model_impl.get_output_embeddings()
|
| 175 |
+
|
| 176 |
+
def set_output_embeddings(self, value):
|
| 177 |
+
self.model_impl.set_output_embeddings(value)
|
| 178 |
+
|
| 179 |
+
def prepare_inputs_for_generation(self, *args, **kwargs):
|
| 180 |
+
return self.model_impl.prepare_inputs_for_generation(*args, **kwargs)
|
| 181 |
+
|
| 182 |
+
@classmethod
|
| 183 |
+
def from_pretrained(cls, pretrained_model_name_or_path, *args, **kwargs):
|
| 184 |
+
"""Load a PMRE model from a HuggingFace repo or local directory.
|
| 185 |
+
|
| 186 |
+
This overrides from_pretrained to:
|
| 187 |
+
1. Load the config (which contains pmre_metadata)
|
| 188 |
+
2. Instantiate with factorized modules
|
| 189 |
+
3. Load the factorized state dict
|
| 190 |
+
"""
|
| 191 |
+
model_dir = Path(pretrained_model_name_or_path)
|
| 192 |
+
|
| 193 |
+
# Check for local factorized_modules.json
|
| 194 |
+
metadata_path = model_dir / "factorized_modules.json" if model_dir.is_dir() else None
|
| 195 |
+
|
| 196 |
+
config = kwargs.pop("config", None)
|
| 197 |
+
if config is None:
|
| 198 |
+
config = AutoConfig.from_pretrained(pretrained_model_name_or_path, trust_remote_code=True)
|
| 199 |
+
|
| 200 |
+
# Inject metadata into config if available from file
|
| 201 |
+
if metadata_path and metadata_path.exists() and not getattr(config, "pmre_metadata", None):
|
| 202 |
+
with metadata_path.open("r", encoding="utf-8") as f:
|
| 203 |
+
config.pmre_metadata = json.load(f)
|
| 204 |
+
|
| 205 |
+
# Build the model with factorized modules
|
| 206 |
+
torch_dtype = kwargs.pop("torch_dtype", None)
|
| 207 |
+
device_map = kwargs.pop("device_map", None)
|
| 208 |
+
|
| 209 |
+
model = cls(config, torch_dtype=torch_dtype)
|
| 210 |
+
|
| 211 |
+
# Load state dict
|
| 212 |
+
from safetensors import safe_open
|
| 213 |
+
from safetensors.torch import load_file
|
| 214 |
+
|
| 215 |
+
safetensors_files = list(model_dir.glob("*.safetensors")) if model_dir.is_dir() else []
|
| 216 |
+
|
| 217 |
+
if safetensors_files:
|
| 218 |
+
state_dict = {}
|
| 219 |
+
for sf_path in safetensors_files:
|
| 220 |
+
state_dict.update(load_file(str(sf_path)))
|
| 221 |
+
# Keys in safetensors are for model_impl
|
| 222 |
+
missing, unexpected = model.model_impl.load_state_dict(state_dict, strict=False)
|
| 223 |
+
if missing:
|
| 224 |
+
print(f" PMRE load: {len(missing)} missing keys (expected for non-factorized buffers)")
|
| 225 |
+
else:
|
| 226 |
+
# Fallback: try .pt/.bin files
|
| 227 |
+
pt_files = list(model_dir.glob("pmre_model.pt")) + list(model_dir.glob("*.bin"))
|
| 228 |
+
if pt_files:
|
| 229 |
+
checkpoint = torch.load(pt_files[0], map_location="cpu", weights_only=False)
|
| 230 |
+
sd = checkpoint.get("model_state_dict", checkpoint)
|
| 231 |
+
model.model_impl.load_state_dict(sd, strict=False)
|
| 232 |
+
|
| 233 |
+
if torch_dtype is not None:
|
| 234 |
+
model = model.to(dtype=torch_dtype)
|
| 235 |
+
if device_map:
|
| 236 |
+
from accelerate import dispatch_model, infer_auto_device_map
|
| 237 |
+
device_map_computed = infer_auto_device_map(model, max_memory=None)
|
| 238 |
+
model = dispatch_model(model, device_map=device_map_computed)
|
| 239 |
+
|
| 240 |
+
model.eval()
|
| 241 |
+
return model
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "<|endoftext|>",
|
| 5 |
+
"clean_up_tokenization_spaces": false,
|
| 6 |
+
"eos_token": "<|endoftext|>",
|
| 7 |
+
"is_local": false,
|
| 8 |
+
"local_files_only": false,
|
| 9 |
+
"model_max_length": 65536,
|
| 10 |
+
"pad_token": "<|pad|>",
|
| 11 |
+
"unk_token": "<|endoftext|>"
|
| 12 |
+
}
|