Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
library_name: transformers
|
| 4 |
+
pipeline_tag: text-generation
|
| 5 |
+
language:
|
| 6 |
+
- bg
|
| 7 |
+
- ca
|
| 8 |
+
- code
|
| 9 |
+
- cs
|
| 10 |
+
- cy
|
| 11 |
+
- da
|
| 12 |
+
- de
|
| 13 |
+
- el
|
| 14 |
+
- en
|
| 15 |
+
- es
|
| 16 |
+
- et
|
| 17 |
+
- eu
|
| 18 |
+
- fi
|
| 19 |
+
- fr
|
| 20 |
+
- ga
|
| 21 |
+
- gl
|
| 22 |
+
- hr
|
| 23 |
+
- hu
|
| 24 |
+
- it
|
| 25 |
+
- lt
|
| 26 |
+
- lv
|
| 27 |
+
- mt
|
| 28 |
+
- nl
|
| 29 |
+
- nn
|
| 30 |
+
- \no
|
| 31 |
+
- oc
|
| 32 |
+
- pl
|
| 33 |
+
- pt
|
| 34 |
+
- ro
|
| 35 |
+
- ru
|
| 36 |
+
- sh
|
| 37 |
+
- sk
|
| 38 |
+
- sl
|
| 39 |
+
- sr
|
| 40 |
+
- sv
|
| 41 |
+
- uk
|
| 42 |
+
datasets:
|
| 43 |
+
- oscar-corpus/colossal-oscar-1.0
|
| 44 |
+
- HuggingFaceFW/fineweb-edu
|
| 45 |
+
- joelniklaus/eurlex_resources
|
| 46 |
+
- joelito/legal-mc4
|
| 47 |
+
- projecte-aina/CATalog
|
| 48 |
+
- UFRGS/brwac
|
| 49 |
+
- community-datasets/hrwac
|
| 50 |
+
- danish-foundation-models/danish-gigaword
|
| 51 |
+
- HiTZ/euscrawl
|
| 52 |
+
- PleIAs/French-PD-Newspapers
|
| 53 |
+
- PleIAs/French-PD-Books
|
| 54 |
+
- AI-team-UoA/greek_legal_code
|
| 55 |
+
- HiTZ/latxa-corpus-v1.1
|
| 56 |
+
- allenai/peS2o
|
| 57 |
+
- pile-of-law/pile-of-law
|
| 58 |
+
- PORTULAN/parlamento-pt
|
| 59 |
+
- hoskinson-center/proof-pile
|
| 60 |
+
- togethercomputer/RedPajama-Data-1T
|
| 61 |
+
- bigcode/starcoderdata
|
| 62 |
+
- bjoernp/tagesschau-2018-2023
|
| 63 |
+
- EleutherAI/the_pile_deduplicated
|
| 64 |
+
base_model: BSC-LT/salamandra-7b-instruct
|
| 65 |
+
tags:
|
| 66 |
+
- mlx
|
| 67 |
+
---
|
| 68 |
+
|
| 69 |
+
# bibproj/salamandra-7b-instruct-mlx-fp16
|
| 70 |
+
|
| 71 |
+
The Model [bibproj/salamandra-7b-instruct-mlx-fp16](https://huggingface.co/bibproj/salamandra-7b-instruct-mlx-fp16) was converted to MLX format from [BSC-LT/salamandra-7b-instruct](https://huggingface.co/BSC-LT/salamandra-7b-instruct) using mlx-lm version **0.22.3**.
|
| 72 |
+
|
| 73 |
+
## Use with mlx
|
| 74 |
+
|
| 75 |
+
```bash
|
| 76 |
+
pip install mlx-lm
|
| 77 |
+
```
|
| 78 |
+
|
| 79 |
+
```python
|
| 80 |
+
from mlx_lm import load, generate
|
| 81 |
+
|
| 82 |
+
model, tokenizer = load("bibproj/salamandra-7b-instruct-mlx-fp16")
|
| 83 |
+
|
| 84 |
+
prompt="hello"
|
| 85 |
+
|
| 86 |
+
if hasattr(tokenizer, "apply_chat_template") and tokenizer.chat_template is not None:
|
| 87 |
+
messages = [{"role": "user", "content": prompt}]
|
| 88 |
+
prompt = tokenizer.apply_chat_template(
|
| 89 |
+
messages, tokenize=False, add_generation_prompt=True
|
| 90 |
+
)
|
| 91 |
+
|
| 92 |
+
response = generate(model, tokenizer, prompt=prompt, verbose=True)
|
| 93 |
+
```
|