Training in progress, step 400

axolotl_config.yaml

@@ -1,4 +1,4 @@
-base_model: fxmarty/tiny-dummy-qwen2
+base_model: peft-internal-testing/tiny-dummy-qwen2
 batch_size: 32
 bf16: true
 chat_template: tokenizer_default_fallback_alpaca
@@ -13,7 +13,9 @@ datasets:
     no_input_format: '{instruction}'
     system_format: '{system}'
     system_prompt: ''
-eval_steps: 20
+device_map: auto
+eval_sample_packing: false
+eval_steps: 200
 flash_attention: true
 gpu_memory_limit: 80GiB
 group_by_length: true
@@ -23,22 +25,24 @@ learning_rate: 0.0002
 logging_steps: 10
 lr_scheduler: cosine
 max_steps: 2500
-micro_batch_size: 2
+micro_batch_size: 4
 model_type: AutoModelForCausalLM
 num_epochs: 100
 optimizer: adamw_bnb_8bit
 output_dir: /root/.sn56/axolotl/outputs/tiny-dummy-qwen2
 pad_to_sequence_len: true
 resize_token_embeddings_to_32x: false
-save_steps: 40
+sample_packing: false
+save_steps: 400
 save_total_limit: 1
 sequence_len: 2048
 tokenizer_type: Qwen2TokenizerFast
+torch_dtype: bf16
 trust_remote_code: true
 val_set_size: 0.1
 wandb_entity: ''
 wandb_mode: online
-wandb_name: fxmarty/tiny-dummy-qwen2-argilla/databricks-dolly-15k-curated-en
+wandb_name: peft-internal-testing/tiny-dummy-qwen2-argilla/databricks-dolly-15k-curated-en
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: default

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "fxmarty/tiny-dummy-qwen2",
+  "_name_or_path": "peft-internal-testing/tiny-dummy-qwen2",
   "architectures": [
     "Qwen2ForCausalLM"
   ],

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:89fe48d90293c6de8a412c54d1431a33bdd6255c0ce7cf015902d4131b92c913
+oid sha256:d74a2d30e9b5eaccdc8860d90690f3f6102f6de30d1c8b6b8ec10411ff53d836
 size 2433024

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:20af897fff432dad9da1232d2f5e6ab4040448b601443dcf983c39a3e47a3b26
-size 6648
+oid sha256:6cafd6770e4d3416c27e43a8308e52009a8c0c1fff65ee59e17f0872cd1b39e9
+size 6840