Commit
•
0f701f1
1
Parent(s):
e492092
Model save
Browse files
README.md
ADDED
|
@@ -0,0 +1,90 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: meta-llama/Meta-Llama-3-8B
|
| 3 |
+
datasets:
|
| 4 |
+
- generator
|
| 5 |
+
library_name: peft
|
| 6 |
+
license: llama3
|
| 7 |
+
tags:
|
| 8 |
+
- trl
|
| 9 |
+
- sft
|
| 10 |
+
- generated_from_trainer
|
| 11 |
+
model-index:
|
| 12 |
+
- name: POC-NEW-Meta-Llama-3-8B-MEDAL-flash-attention-2-cosine-evaldata
|
| 13 |
+
results: []
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 17 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 18 |
+
|
| 19 |
+
# POC-NEW-Meta-Llama-3-8B-MEDAL-flash-attention-2-cosine-evaldata
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B) on the generator dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 2.2356
|
| 24 |
+
|
| 25 |
+
## Model description
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Intended uses & limitations
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Training and evaluation data
|
| 34 |
+
|
| 35 |
+
More information needed
|
| 36 |
+
|
| 37 |
+
## Training procedure
|
| 38 |
+
|
| 39 |
+
### Training hyperparameters
|
| 40 |
+
|
| 41 |
+
The following hyperparameters were used during training:
|
| 42 |
+
- learning_rate: 0.0002
|
| 43 |
+
- train_batch_size: 3
|
| 44 |
+
- eval_batch_size: 8
|
| 45 |
+
- seed: 42
|
| 46 |
+
- gradient_accumulation_steps: 8
|
| 47 |
+
- total_train_batch_size: 24
|
| 48 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 49 |
+
- lr_scheduler_type: constant
|
| 50 |
+
- lr_scheduler_warmup_ratio: 0.03
|
| 51 |
+
- lr_scheduler_warmup_steps: 1500
|
| 52 |
+
- num_epochs: 0.5
|
| 53 |
+
|
| 54 |
+
### Training results
|
| 55 |
+
|
| 56 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 57 |
+
|:-------------:|:------:|:----:|:---------------:|
|
| 58 |
+
| 2.4484 | 0.0207 | 100 | 2.3720 |
|
| 59 |
+
| 2.3535 | 0.0415 | 200 | 2.3370 |
|
| 60 |
+
| 2.3303 | 0.0622 | 300 | 2.3204 |
|
| 61 |
+
| 2.3153 | 0.0830 | 400 | 2.3081 |
|
| 62 |
+
| 2.3041 | 0.1037 | 500 | 2.2982 |
|
| 63 |
+
| 2.2904 | 0.1245 | 600 | 2.2917 |
|
| 64 |
+
| 2.2954 | 0.1452 | 700 | 2.2845 |
|
| 65 |
+
| 2.2795 | 0.1660 | 800 | 2.2790 |
|
| 66 |
+
| 2.2772 | 0.1867 | 900 | 2.2751 |
|
| 67 |
+
| 2.2769 | 0.2075 | 1000 | 2.2711 |
|
| 68 |
+
| 2.2711 | 0.2282 | 1100 | 2.2678 |
|
| 69 |
+
| 2.2722 | 0.2489 | 1200 | 2.2644 |
|
| 70 |
+
| 2.269 | 0.2697 | 1300 | 2.2610 |
|
| 71 |
+
| 2.2651 | 0.2904 | 1400 | 2.2586 |
|
| 72 |
+
| 2.2625 | 0.3112 | 1500 | 2.2550 |
|
| 73 |
+
| 2.2579 | 0.3319 | 1600 | 2.2516 |
|
| 74 |
+
| 2.2532 | 0.3527 | 1700 | 2.2501 |
|
| 75 |
+
| 2.256 | 0.3734 | 1800 | 2.2471 |
|
| 76 |
+
| 2.2509 | 0.3942 | 1900 | 2.2450 |
|
| 77 |
+
| 2.2482 | 0.4149 | 2000 | 2.2433 |
|
| 78 |
+
| 2.247 | 0.4357 | 2100 | 2.2406 |
|
| 79 |
+
| 2.2404 | 0.4564 | 2200 | 2.2395 |
|
| 80 |
+
| 2.2377 | 0.4771 | 2300 | 2.2372 |
|
| 81 |
+
| 2.2373 | 0.4979 | 2400 | 2.2356 |
|
| 82 |
+
|
| 83 |
+
|
| 84 |
+
### Framework versions
|
| 85 |
+
|
| 86 |
+
- PEFT 0.11.1
|
| 87 |
+
- Transformers 4.41.2
|
| 88 |
+
- Pytorch 2.3.0+cu121
|
| 89 |
+
- Datasets 2.20.0
|
| 90 |
+
- Tokenizers 0.19.1
|
logs/events.out.tfevents.1719190919.1fa3d88adcea.55250.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:154c98af9e2d4097f32295d3678cf831c3fc05c450a806cae35f3bfcd91b066d
|
| 3 |
+
size 17522
|