End of training
Browse files- README.md +13 -8
- adapter_model.bin +1 -1
README.md
CHANGED
|
@@ -34,7 +34,7 @@ datasets:
|
|
| 34 |
- path: ft_data/alpaca_data.jsonl
|
| 35 |
type: alpaca
|
| 36 |
dataset_prepared_path: last_run_prepared
|
| 37 |
-
val_set_size: 0.
|
| 38 |
output_dir: ./qlora-alpaca-out
|
| 39 |
hub_model_id: pbevan11/llama-3.1-8b-ocr-correction
|
| 40 |
|
|
@@ -43,7 +43,6 @@ lora_model_dir:
|
|
| 43 |
|
| 44 |
sequence_len: 8192
|
| 45 |
sample_packing: true
|
| 46 |
-
eval_sample_packing: false
|
| 47 |
pad_to_sequence_len: true
|
| 48 |
|
| 49 |
lora_r: 32
|
|
@@ -104,12 +103,12 @@ special_tokens:
|
|
| 104 |
|
| 105 |
</details><br>
|
| 106 |
|
| 107 |
-
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/sncds/ocr-ft/runs/
|
| 108 |
# llama-3.1-8b-ocr-correction
|
| 109 |
|
| 110 |
This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B) on the None dataset.
|
| 111 |
It achieves the following results on the evaluation set:
|
| 112 |
-
- Loss: 0.
|
| 113 |
|
| 114 |
## Model description
|
| 115 |
|
|
@@ -141,10 +140,16 @@ The following hyperparameters were used during training:
|
|
| 141 |
|
| 142 |
### Training results
|
| 143 |
|
| 144 |
-
| Training Loss | Epoch
|
| 145 |
-
|
| 146 |
-
| 0.
|
| 147 |
-
| 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 148 |
|
| 149 |
|
| 150 |
### Framework versions
|
|
|
|
| 34 |
- path: ft_data/alpaca_data.jsonl
|
| 35 |
type: alpaca
|
| 36 |
dataset_prepared_path: last_run_prepared
|
| 37 |
+
val_set_size: 0.05
|
| 38 |
output_dir: ./qlora-alpaca-out
|
| 39 |
hub_model_id: pbevan11/llama-3.1-8b-ocr-correction
|
| 40 |
|
|
|
|
| 43 |
|
| 44 |
sequence_len: 8192
|
| 45 |
sample_packing: true
|
|
|
|
| 46 |
pad_to_sequence_len: true
|
| 47 |
|
| 48 |
lora_r: 32
|
|
|
|
| 103 |
|
| 104 |
</details><br>
|
| 105 |
|
| 106 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/sncds/ocr-ft/runs/rotjhntf)
|
| 107 |
# llama-3.1-8b-ocr-correction
|
| 108 |
|
| 109 |
This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B) on the None dataset.
|
| 110 |
It achieves the following results on the evaluation set:
|
| 111 |
+
- Loss: 0.1901
|
| 112 |
|
| 113 |
## Model description
|
| 114 |
|
|
|
|
| 140 |
|
| 141 |
### Training results
|
| 142 |
|
| 143 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 144 |
+
|:-------------:|:------:|:----:|:---------------:|
|
| 145 |
+
| 0.61 | 0.0331 | 1 | 0.6018 |
|
| 146 |
+
| 0.4379 | 0.2645 | 8 | 0.4256 |
|
| 147 |
+
| 0.2531 | 0.5289 | 16 | 0.2714 |
|
| 148 |
+
| 0.2366 | 0.7934 | 24 | 0.2247 |
|
| 149 |
+
| 0.1839 | 1.0331 | 32 | 0.2053 |
|
| 150 |
+
| 0.1752 | 1.2975 | 40 | 0.1961 |
|
| 151 |
+
| 0.1629 | 1.5620 | 48 | 0.1909 |
|
| 152 |
+
| 0.163 | 1.8264 | 56 | 0.1901 |
|
| 153 |
|
| 154 |
|
| 155 |
### Framework versions
|
adapter_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 167934026
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:befe7ee91cb8ab62450880c1dabf645b053b56d4e5b4cf5a4776e29329224eeb
|
| 3 |
size 167934026
|