Commit
•
87bef75
1
Parent(s):
d90f517
Model save
Browse files
README.md
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
library_name: peft
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- sft
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
datasets:
|
| 9 |
+
- generator
|
| 10 |
+
base_model: mistralai/Mixtral-8x7B-v0.1
|
| 11 |
+
model-index:
|
| 12 |
+
- name: mixtral_mix_2
|
| 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 |
+
# mixtral_mix_2
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [mistralai/Mixtral-8x7B-v0.1](https://huggingface.co/mistralai/Mixtral-8x7B-v0.1) on the generator dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 0.2832
|
| 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: 0.0001
|
| 43 |
+
- train_batch_size: 2
|
| 44 |
+
- eval_batch_size: 1
|
| 45 |
+
- seed: 42
|
| 46 |
+
- gradient_accumulation_steps: 32
|
| 47 |
+
- total_train_batch_size: 64
|
| 48 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 49 |
+
- lr_scheduler_type: cosine
|
| 50 |
+
- lr_scheduler_warmup_ratio: 0.03
|
| 51 |
+
- training_steps: 200
|
| 52 |
+
|
| 53 |
+
### Training results
|
| 54 |
+
|
| 55 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 56 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
| 57 |
+
| 1.686 | 0.36 | 10 | 1.1952 |
|
| 58 |
+
| 1.0961 | 0.71 | 20 | 0.9786 |
|
| 59 |
+
| 0.9359 | 1.07 | 30 | 0.8352 |
|
| 60 |
+
| 0.7603 | 1.42 | 40 | 0.6454 |
|
| 61 |
+
| 0.5832 | 1.78 | 50 | 0.4951 |
|
| 62 |
+
| 0.4764 | 2.14 | 60 | 0.4361 |
|
| 63 |
+
| 0.4172 | 2.49 | 70 | 0.3998 |
|
| 64 |
+
| 0.3911 | 2.85 | 80 | 0.3735 |
|
| 65 |
+
| 0.37 | 3.2 | 90 | 0.3564 |
|
| 66 |
+
| 0.3408 | 3.56 | 100 | 0.3407 |
|
| 67 |
+
| 0.3454 | 3.92 | 110 | 0.3270 |
|
| 68 |
+
| 0.3153 | 4.27 | 120 | 0.3142 |
|
| 69 |
+
| 0.3025 | 4.63 | 130 | 0.3046 |
|
| 70 |
+
| 0.3076 | 4.98 | 140 | 0.2963 |
|
| 71 |
+
| 0.2791 | 5.34 | 150 | 0.2903 |
|
| 72 |
+
| 0.2907 | 5.7 | 160 | 0.2863 |
|
| 73 |
+
| 0.2787 | 6.05 | 170 | 0.2840 |
|
| 74 |
+
| 0.284 | 6.41 | 180 | 0.2832 |
|
| 75 |
+
| 0.264 | 6.76 | 190 | 0.2832 |
|
| 76 |
+
| 0.2734 | 7.12 | 200 | 0.2832 |
|
| 77 |
+
|
| 78 |
+
|
| 79 |
+
### Framework versions
|
| 80 |
+
|
| 81 |
+
- PEFT 0.7.2.dev0
|
| 82 |
+
- Transformers 4.38.1
|
| 83 |
+
- Pytorch 2.1.2+cu121
|
| 84 |
+
- Datasets 2.16.1
|
| 85 |
+
- Tokenizers 0.15.0
|