qwen-final-databricks/databricks-dolly-15k
Browse files
README.md
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: Qwen/Qwen-14B
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
model-index:
|
| 6 |
+
- name: final_databricks_databricks-dolly-15k
|
| 7 |
+
results: []
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 11 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 12 |
+
|
| 13 |
+
# final_databricks_databricks-dolly-15k
|
| 14 |
+
|
| 15 |
+
This model is a fine-tuned version of [Qwen/Qwen-14B](https://huggingface.co/Qwen/Qwen-14B) on an unknown dataset.
|
| 16 |
+
It achieves the following results on the evaluation set:
|
| 17 |
+
- Loss: 1.6083
|
| 18 |
+
|
| 19 |
+
## Model description
|
| 20 |
+
|
| 21 |
+
More information needed
|
| 22 |
+
|
| 23 |
+
## Intended uses & limitations
|
| 24 |
+
|
| 25 |
+
More information needed
|
| 26 |
+
|
| 27 |
+
## Training and evaluation data
|
| 28 |
+
|
| 29 |
+
More information needed
|
| 30 |
+
|
| 31 |
+
## Training procedure
|
| 32 |
+
|
| 33 |
+
### Training hyperparameters
|
| 34 |
+
|
| 35 |
+
The following hyperparameters were used during training:
|
| 36 |
+
- learning_rate: 0.0003
|
| 37 |
+
- train_batch_size: 2
|
| 38 |
+
- eval_batch_size: 2
|
| 39 |
+
- seed: 42
|
| 40 |
+
- gradient_accumulation_steps: 132
|
| 41 |
+
- total_train_batch_size: 264
|
| 42 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 43 |
+
- lr_scheduler_type: cosine
|
| 44 |
+
- lr_scheduler_warmup_steps: 0.01
|
| 45 |
+
- num_epochs: 1
|
| 46 |
+
|
| 47 |
+
### Training results
|
| 48 |
+
|
| 49 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 50 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
| 51 |
+
| 1.694 | 0.04 | 2 | 1.8016 |
|
| 52 |
+
| 1.6398 | 0.07 | 4 | 1.7369 |
|
| 53 |
+
| 1.6421 | 0.11 | 6 | 1.6886 |
|
| 54 |
+
| 1.579 | 0.15 | 8 | 1.6596 |
|
| 55 |
+
| 1.5589 | 0.18 | 10 | 1.6420 |
|
| 56 |
+
| 1.5944 | 0.22 | 12 | 1.6305 |
|
| 57 |
+
| 1.5314 | 0.26 | 14 | 1.6274 |
|
| 58 |
+
| 1.5841 | 0.29 | 16 | 1.6238 |
|
| 59 |
+
| 1.5945 | 0.33 | 18 | 1.6229 |
|
| 60 |
+
| 1.5755 | 0.37 | 20 | 1.6234 |
|
| 61 |
+
| 1.5527 | 0.4 | 22 | 1.6231 |
|
| 62 |
+
| 1.6121 | 0.44 | 24 | 1.6224 |
|
| 63 |
+
| 1.586 | 0.48 | 26 | 1.6219 |
|
| 64 |
+
| 1.5995 | 0.52 | 28 | 1.6213 |
|
| 65 |
+
| 1.5942 | 0.55 | 30 | 1.6200 |
|
| 66 |
+
| 1.5738 | 0.59 | 32 | 1.6180 |
|
| 67 |
+
| 1.5825 | 0.63 | 34 | 1.6161 |
|
| 68 |
+
| 1.5183 | 0.66 | 36 | 1.6137 |
|
| 69 |
+
| 1.5964 | 0.7 | 38 | 1.6120 |
|
| 70 |
+
| 1.623 | 0.74 | 40 | 1.6105 |
|
| 71 |
+
| 1.5783 | 0.77 | 42 | 1.6098 |
|
| 72 |
+
| 1.6046 | 0.81 | 44 | 1.6093 |
|
| 73 |
+
| 1.5157 | 0.85 | 46 | 1.6088 |
|
| 74 |
+
| 1.5317 | 0.88 | 48 | 1.6086 |
|
| 75 |
+
| 1.5578 | 0.92 | 50 | 1.6086 |
|
| 76 |
+
| 1.5402 | 0.96 | 52 | 1.6084 |
|
| 77 |
+
| 1.5616 | 0.99 | 54 | 1.6083 |
|
| 78 |
+
|
| 79 |
+
|
| 80 |
+
### Framework versions
|
| 81 |
+
|
| 82 |
+
- Transformers 4.32.0
|
| 83 |
+
- Pytorch 2.1.0
|
| 84 |
+
- Datasets 2.14.7
|
| 85 |
+
- Tokenizers 0.13.3
|