Update README.md
Browse files
README.md
CHANGED
@@ -20,10 +20,64 @@ should probably proofread and complete it, then remove this comment. -->
|
|
20 |
|
21 |
This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) on the generator dataset.
|
22 |
|
|
|
|
|
23 |
## Model description
|
24 |
|
25 |
More information needed
|
26 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
27 |
## Intended uses & limitations
|
28 |
|
29 |
More information needed
|
|
|
20 |
|
21 |
This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) on the generator dataset.
|
22 |
|
23 |
+
with dataset b-mc2/sql-create-context
|
24 |
+
|
25 |
## Model description
|
26 |
|
27 |
More information needed
|
28 |
|
29 |
+
### Testing results
|
30 |
+
|
31 |
+
import torch
|
32 |
+
from peft import AutoPeftModelForCausalLM
|
33 |
+
from transformers import AutoTokenizer, pipeline
|
34 |
+
|
35 |
+
peft_model_id = "frankmorales2020/Mistral-7B-text-to-sql-without-flash-attention-2"
|
36 |
+
|
37 |
+
# Load Model with PEFT adapter
|
38 |
+
model = AutoPeftModelForCausalLM.from_pretrained(
|
39 |
+
peft_model_id,
|
40 |
+
device_map="auto",
|
41 |
+
torch_dtype=torch.float16
|
42 |
+
)
|
43 |
+
|
44 |
+
tokenizer = AutoTokenizer.from_pretrained(peft_model_id)
|
45 |
+
|
46 |
+
# load into pipeline
|
47 |
+
pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)
|
48 |
+
|
49 |
+
# CASE Number 1:
|
50 |
+
prompt='What was the first album Beyoncé released as a solo artist?'
|
51 |
+
prompt = f"Instruct: generate a SQL query.\n{prompt}\nOutput:\n" # for dataset b-mc2/sql-create-context
|
52 |
+
outputs = pipe(prompt, max_new_tokens=1024, do_sample=True, temperature=0.9, top_k=50, top_p=0.1, eos_token_id=pipe.tokenizer.eos_token_id, pad_token_id=pipe.tokenizer.eos_token_id)
|
53 |
+
|
54 |
+
print('Question: %s'%prompt)
|
55 |
+
#print('Answer: %s \nOutput:\n'%outputs[0]['generated_text'])
|
56 |
+
print(f"Generated Answer:\n{outputs[0]['generated_text'][len(prompt):].strip()}")
|
57 |
+
|
58 |
+
Question: Instruct: generate a SQL query.
|
59 |
+
What was the first album Beyoncé released as a solo artist?
|
60 |
+
Output:
|
61 |
+
|
62 |
+
Generated Answer:
|
63 |
+
SELECT first_album FROM table_name_82 WHERE solo_artist = "beyoncé"
|
64 |
+
|
65 |
+
# CASE Number 2:
|
66 |
+
prompt='What was the first album Beyoncé released as a solo artist?'
|
67 |
+
prompt = f"Instruct: Answer the following question.\n{prompt}\nOutput:\n"
|
68 |
+
outputs = pipe(prompt, max_new_tokens=1024, do_sample=True, temperature=0.9, top_k=50, top_p=0.1, eos_token_id=pipe.tokenizer.eos_token_id, pad_token_id=pipe.tokenizer.eos_token_id)
|
69 |
+
|
70 |
+
print('Question: %s'%prompt)
|
71 |
+
#print('Answer: %s \nOutput:\n'%outputs[0]['generated_text'])
|
72 |
+
print(f"Generated Answer:\n{outputs[0]['generated_text'][len(prompt):].strip()}")
|
73 |
+
|
74 |
+
Question: Instruct: Answer the following question.
|
75 |
+
What was the first album Beyoncé released as a solo artist?
|
76 |
+
Output:
|
77 |
+
|
78 |
+
Generated Answer:
|
79 |
+
The first album Beyoncé released as a solo artist was "Dangerously in Love".
|
80 |
+
|
81 |
## Intended uses & limitations
|
82 |
|
83 |
More information needed
|