Update main branch model card with loading instructions.
Browse files
README.md
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
---
|
| 2 |
-
version:
|
| 3 |
family: smollm2-360m
|
| 4 |
model_name: smollm2-360m-all_raw_folders_baseline-300B-mbs16-gbs1024-16feb-lr2e-05-gbs16
|
| 5 |
license: mit
|
|
@@ -8,7 +8,7 @@ tags:
|
|
| 8 |
- transformer
|
| 9 |
- smollm2
|
| 10 |
---
|
| 11 |
-
# SmolLM2 smollm2-360m-all_raw_folders_baseline-300B-mbs16-gbs1024-16feb-lr2e-05-gbs16 (Version:
|
| 12 |
|
| 13 |
## Model Details
|
| 14 |
- **Architecture:** SmolLM2
|
|
@@ -44,3 +44,18 @@ train:
|
|
| 44 |
tie_embeddings: false
|
| 45 |
|
| 46 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
version: main
|
| 3 |
family: smollm2-360m
|
| 4 |
model_name: smollm2-360m-all_raw_folders_baseline-300B-mbs16-gbs1024-16feb-lr2e-05-gbs16
|
| 5 |
license: mit
|
|
|
|
| 8 |
- transformer
|
| 9 |
- smollm2
|
| 10 |
---
|
| 11 |
+
# SmolLM2 smollm2-360m-all_raw_folders_baseline-300B-mbs16-gbs1024-16feb-lr2e-05-gbs16 (Version: main)
|
| 12 |
|
| 13 |
## Model Details
|
| 14 |
- **Architecture:** SmolLM2
|
|
|
|
| 44 |
tie_embeddings: false
|
| 45 |
|
| 46 |
```
|
| 47 |
+
|
| 48 |
+
|
| 49 |
+
## Model Loading and Revision System
|
| 50 |
+
|
| 51 |
+
This repository hosts multiple revisions of the model.
|
| 52 |
+
To load a specific revision, use the `revision` parameter. For example:
|
| 53 |
+
|
| 54 |
+
```python
|
| 55 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 56 |
+
|
| 57 |
+
model = AutoModelForCausalLM.from_pretrained("locuslab/mix_ift_v3-smollm2-360m-smollm2-360m-all_raw_folders_baseline", revision="final")
|
| 58 |
+
tokenizer = AutoTokenizer.from_pretrained("locuslab/mix_ift_v3-smollm2-360m-smollm2-360m-all_raw_folders_baseline", revision="final")
|
| 59 |
+
```
|
| 60 |
+
|
| 61 |
+
Replace `"final"` with the desired revision.
|