pankajmathur
commited on
Commit
•
665bccf
1
Parent(s):
befabf7
Update README.md
Browse files
README.md
CHANGED
@@ -8,11 +8,11 @@ pipeline_tag: text2text-generation
|
|
8 |
|
9 |
**Model Name: llama_3_orca_mini_v4_8b**
|
10 |
|
11 |
-
|
12 |
|
13 |
-
<img src="" width="
|
14 |
|
15 |
-
NOTICE
|
16 |
|
17 |
By providing proper credit and attribution, you are granted permission to use this model as a foundational base for further DPO/PPO tuning or Merges.
|
18 |
I actively encourage users to customize and enhance the model according to their specific needs, as this version is designed to be a comprehensive, fully fine-tuned general model.
|
@@ -22,17 +22,17 @@ Dive in and innovate!
|
|
22 |
|
23 |
We evaluated this model on a wide range of tasks using [Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness) from EleutherAI.
|
24 |
|
25 |
-
Here are the results on metrics used by [HuggingFaceH4 Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
26 |
|
27 |
| Metric |Value|
|
28 |
|---------------------------------|----:|
|
29 |
-
|Avg. |
|
30 |
-
|AI2 Reasoning Challenge (25-Shot)|
|
31 |
-
|HellaSwag (10-Shot) |
|
32 |
-
|MMLU (5-Shot) |
|
33 |
-
|TruthfulQA (0-shot) |
|
34 |
-
|Winogrande (5-shot) |
|
35 |
-
|GSM8k (5-shot) |
|
36 |
|
37 |
|
38 |
<br>
|
@@ -52,17 +52,17 @@ Hello Orca Mini, what can you do for me?<|im_end|>
|
|
52 |
Below shows a code example on how to use this model
|
53 |
|
54 |
```python
|
55 |
-
from transformers import
|
56 |
-
|
57 |
-
model = AutoModel.from_pretrained(
|
58 |
-
tokenizer = AutoTokenizer.from_pretrained(
|
59 |
|
60 |
messages = [
|
61 |
{"role": "system", "content": "You are Orca Mini, a helpful AI assistant."},
|
62 |
{"role": "user", "content": "Hello Orca Mini, what can you do for me?"}
|
63 |
]
|
64 |
|
65 |
-
gen_input = tokenizer.apply_chat_template(
|
66 |
model.generate(**gen_input)
|
67 |
```
|
68 |
This model is governed by [META LLAMA 3 COMMUNITY LICENSE AGREEMENT](LICENSE)
|
|
|
8 |
|
9 |
**Model Name: llama_3_orca_mini_v4_8b**
|
10 |
|
11 |
+
# Llama-3-8b base model trained on Orca Style Mini Datasets
|
12 |
|
13 |
+
<img src="https://huggingface.co/pankajmathur/orca_mini_v4_8b/resolve/main/orca_minis_small.jpeg" width="auto" />
|
14 |
|
15 |
+
## NOTICE
|
16 |
|
17 |
By providing proper credit and attribution, you are granted permission to use this model as a foundational base for further DPO/PPO tuning or Merges.
|
18 |
I actively encourage users to customize and enhance the model according to their specific needs, as this version is designed to be a comprehensive, fully fine-tuned general model.
|
|
|
22 |
|
23 |
We evaluated this model on a wide range of tasks using [Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness) from EleutherAI.
|
24 |
|
25 |
+
Here are the results on similar metrics used by [HuggingFaceH4 Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
26 |
|
27 |
| Metric |Value|
|
28 |
|---------------------------------|----:|
|
29 |
+
|Avg. |67.75|
|
30 |
+
|AI2 Reasoning Challenge (25-Shot)|59.13|
|
31 |
+
|HellaSwag (10-Shot) |81.40|
|
32 |
+
|MMLU (5-Shot) |63.08|
|
33 |
+
|TruthfulQA (0-shot) |55.40|
|
34 |
+
|Winogrande (5-shot) |74.03|
|
35 |
+
|GSM8k (5-shot) |73.46|
|
36 |
|
37 |
|
38 |
<br>
|
|
|
52 |
Below shows a code example on how to use this model
|
53 |
|
54 |
```python
|
55 |
+
from transformers import AutoModel, AutoTokenizer
|
56 |
+
model_slug = "pankajmathur/orca_mini_v4_8b"
|
57 |
+
model = AutoModel.from_pretrained(model_slug)
|
58 |
+
tokenizer = AutoTokenizer.from_pretrained(model_slug)
|
59 |
|
60 |
messages = [
|
61 |
{"role": "system", "content": "You are Orca Mini, a helpful AI assistant."},
|
62 |
{"role": "user", "content": "Hello Orca Mini, what can you do for me?"}
|
63 |
]
|
64 |
|
65 |
+
gen_input = tokenizer.apply_chat_template(messages, return_tensors="pt")
|
66 |
model.generate(**gen_input)
|
67 |
```
|
68 |
This model is governed by [META LLAMA 3 COMMUNITY LICENSE AGREEMENT](LICENSE)
|