Update README.md
Browse files
README.md
CHANGED
@@ -158,7 +158,7 @@ In this example, the model needs to understand and generate a coherent conversat
|
|
158 |
|
159 |
```bash
|
160 |
python sft_14.py \
|
161 |
-
--run_name="llama3.1-8b-
|
162 |
--model_name_or_path="meta-llama/Meta-Llama-3.1-8B" \
|
163 |
--dataset_name="mlfoundations/dclm-baseline-1.0-parquet,mlabonne/FineTome-100k" \
|
164 |
--report_to="wandb" \
|
@@ -177,13 +177,13 @@ python sft_14.py \
|
|
177 |
--logging_steps=1 \
|
178 |
--num_train_epochs=1 \
|
179 |
--push_to_hub \
|
180 |
-
--hub_model_id="ericflo/Llama-3.1-8B-
|
181 |
--hub_strategy="all_checkpoints" \
|
182 |
--gradient_checkpointing \
|
183 |
--use_liger=true \
|
184 |
--packing=true \
|
185 |
--torch_dtype="bfloat16" \
|
186 |
-
--output_dir="
|
187 |
```
|
188 |
|
189 |
## Intended Uses
|
@@ -210,4 +210,4 @@ This model is designed for:
|
|
210 |
|
211 |
## Contact
|
212 |
|
213 |
-
For inquiries about this model, please contact Eric Florenzano through the [model repository](https://huggingface.co/ericflo/Llama-3.1-8B-
|
|
|
158 |
|
159 |
```bash
|
160 |
python sft_14.py \
|
161 |
+
--run_name="llama3.1-8b-continued2" \
|
162 |
--model_name_or_path="meta-llama/Meta-Llama-3.1-8B" \
|
163 |
--dataset_name="mlfoundations/dclm-baseline-1.0-parquet,mlabonne/FineTome-100k" \
|
164 |
--report_to="wandb" \
|
|
|
177 |
--logging_steps=1 \
|
178 |
--num_train_epochs=1 \
|
179 |
--push_to_hub \
|
180 |
+
--hub_model_id="ericflo/Llama-3.1-8B-ContinuedTraining2-FFT" \
|
181 |
--hub_strategy="all_checkpoints" \
|
182 |
--gradient_checkpointing \
|
183 |
--use_liger=true \
|
184 |
--packing=true \
|
185 |
--torch_dtype="bfloat16" \
|
186 |
+
--output_dir="continuedtraining2_output"
|
187 |
```
|
188 |
|
189 |
## Intended Uses
|
|
|
210 |
|
211 |
## Contact
|
212 |
|
213 |
+
For inquiries about this model, please contact Eric Florenzano through the [model repository](https://huggingface.co/ericflo/Llama-3.1-8B-ContinuedTraining2-FFT).
|