Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
- fr
|
| 6 |
+
- de
|
| 7 |
+
- es
|
| 8 |
+
- it
|
| 9 |
+
- pt
|
| 10 |
+
- ru
|
| 11 |
+
- zh
|
| 12 |
+
- ja
|
| 13 |
+
pipeline_tag: text-generation
|
| 14 |
+
tags:
|
| 15 |
+
- chat
|
| 16 |
+
---
|
| 17 |
+
|
| 18 |
+
## This repo contains EXL2 quants of the model. If you need the original weights, please find them [here](https://huggingface.co/anthracite-org/magnum-v2-72b).
|
| 19 |
+
## Base repo only contains the measurement file, see revisions for your quant of choice.
|
| 20 |
+
|
| 21 |
+
- [measurement.json](https://huggingface.co/anthracite-org/magnum-v2-72b-exl2/tree/main)
|
| 22 |
+
- [3.0bpw](https://huggingface.co/anthracite-org/magnum-v2-72b-exl2/tree/2.7bpw)
|
| 23 |
+
- [4.0bpw](https://huggingface.co/anthracite-org/magnum-v2-72b-exl2/tree/4.0bpw)
|
| 24 |
+
- [6.0bpw](https://huggingface.co/anthracite-org/magnum-v2-72b-exl2/tree/6.0bpw)
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
|
| 28 |
+
|
| 29 |
+
This is the seventh (Lucky!) in a series of models designed to replicate the prose quality of the Claude 3 models, specifically Sonnet and Opus. This model is fine-tuned on top of [Qwen-2 72B Instruct](https://huggingface.co/Qwen/Qwen2-72B-Instruct).
|
| 30 |
+
|
| 31 |
+
## Prompting
|
| 32 |
+
Model has been Instruct tuned with the ChatML formatting. A typical input would look like this:
|
| 33 |
+
|
| 34 |
+
```py
|
| 35 |
+
"""<|im_start|>user
|
| 36 |
+
Hi there!<|im_end|>
|
| 37 |
+
<|im_start|>assistant
|
| 38 |
+
Nice to meet you!<|im_end|>
|
| 39 |
+
<|im_start|>user
|
| 40 |
+
Can I ask a question?<|im_end|>
|
| 41 |
+
<|im_start|>assistant
|
| 42 |
+
"""
|
| 43 |
+
```
|
| 44 |
+
|
| 45 |
+
## Credits
|
| 46 |
+
- [anthracite-org/Stheno-Data-Filtered](https://huggingface.co/datasets/anthracite-org/Stheno-Data-Filtered)
|
| 47 |
+
- [anthracite-org/kalo-opus-instruct-22k-no-refusal](https://huggingface.co/datasets/anthracite-org/kalo-opus-instruct-22k-no-refusal)
|
| 48 |
+
- [anthracite-org/nopm_claude_writing_fixed](https://huggingface.co/datasets/anthracite-org/nopm_claude_writing_fixed)
|
| 49 |
+
|
| 50 |
+
This model has been a team effort, and the credits goes to all members of Anthracite.
|
| 51 |
+
|
| 52 |
+
## Training
|
| 53 |
+
The training was done for 2 epochs. We used 8x [AMD Instinct™ MI300X Accelerators](https://www.amd.com/en/products/accelerators/instinct/mi300/mi300x.html) for the full-parameter fine-tuning of the model.
|
| 54 |
+
|
| 55 |
+
We also trained with a weight decay of 0.01 to help further stabilize the loss trajectory and mitigate catastrophic forgetting, and utilize a peak learning rate of 4e-6 to prevent the 2nd epoch loss from dropping too significantly (as it is a strong indicator of overfitting).
|
| 56 |
+
|
| 57 |
+
|
| 58 |
+
Sample Packing was done for 16k tokens rather than the 8k tokens used in our previous runs.
|
| 59 |
+
|
| 60 |
+
[<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
|
| 61 |
+
|
| 62 |
+
## Safety
|
| 63 |
+
...
|