Update README.md
Browse files
README.md
CHANGED
@@ -1,5 +1,37 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
4 |
A Mistral7B Instruct (https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1)
|
5 |
-
Finetune using QLoRA on the docs available in https://docs.modular.com/mojo/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
+
pipeline_tag: text-generation
|
4 |
+
tags:
|
5 |
+
- finetuned
|
6 |
+
inference:
|
7 |
+
parameters:
|
8 |
+
temperature: 0.1
|
9 |
---
|
10 |
+
|
11 |
A Mistral7B Instruct (https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1)
|
12 |
+
Finetune using QLoRA on the docs available in https://docs.modular.com/mojo/
|
13 |
+
|
14 |
+
The Mistral-7B-Instruct-v0.1 Large Language Model (LLM) is a instruct fine-tuned version of the [Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) generative text model using a variety of publicly available conversation datasets.
|
15 |
+
|
16 |
+
For full details of this model please read our [paper](https://arxiv.org/abs/2310.06825) and [release blog post](https://mistral.ai/news/announcing-mistral-7b/).
|
17 |
+
|
18 |
+
## Instruction format
|
19 |
+
```python
|
20 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
21 |
+
|
22 |
+
device = "cuda" # the device to load the model onto
|
23 |
+
|
24 |
+
model = AutoModelForCausalLM.from_pretrained("mcysqrd/MODULARMOJO_Mistral-V1")
|
25 |
+
tokenizer = AutoTokenizer.from_pretrained("mcysqrd/MODULARMOJO_Mistral-V1")
|
26 |
+
|
27 |
+
message = "What can you tell me about MODULAR_MOJO mojo_roadmap Scoping and mutability of statement variables?
|
28 |
+
|
29 |
+
encodeds = tokenizer.apply_chat_template(message, return_tensors="pt")
|
30 |
+
|
31 |
+
model_inputs = encodeds.to(device)
|
32 |
+
model.to(device)
|
33 |
+
|
34 |
+
generated_ids = model.generate(model_inputs, max_new_tokens=1650, do_sample=True, temperature = 0.01)
|
35 |
+
decoded = tokenizer.batch_decode(generated_ids)
|
36 |
+
print(decoded[0])
|
37 |
+
```
|