ayoubkirouane
commited on
Commit
•
7a07883
1
Parent(s):
596068e
Update README.md
Browse files
README.md
CHANGED
@@ -1,6 +1,44 @@
|
|
1 |
---
|
2 |
library_name: peft
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
4 |
## Training procedure
|
5 |
|
6 |
|
@@ -18,3 +56,42 @@ The following `bitsandbytes` quantization config was used during training:
|
|
18 |
|
19 |
|
20 |
- PEFT 0.4.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
library_name: peft
|
3 |
+
license: llama2
|
4 |
+
datasets:
|
5 |
+
- TuningAI/Cover_letter_v2
|
6 |
+
language:
|
7 |
+
- en
|
8 |
+
pipeline_tag: text-generation
|
9 |
---
|
10 |
+
## Model Name: **Llama2_7B_Cover_letter_generator**
|
11 |
+
## Description:
|
12 |
+
**Llama2_7B_Cover_letter_generator** is a powerful, custom language model that has been meticulously fine-tuned to excel at generating cover letters for various job positions.
|
13 |
+
It serves as an invaluable tool for automating the creation of personalized cover letters, tailored to specific job descriptions.
|
14 |
+
## Base Model:
|
15 |
+
This model is based on the Meta's "meta-llama/Llama-2-7b-hf" architecture, making it a highly capable foundation for generating human-like text responses.
|
16 |
+
|
17 |
+
## Dataset :
|
18 |
+
This model was fine-tuned on a custom dataset meticulously curated with more than 200 unique examples.
|
19 |
+
The dataset incorporates both manual entries and contributions from GPT3.5, GPT4, and Falcon 180B models.
|
20 |
+
|
21 |
+
## Fine-tuning Techniques:
|
22 |
+
Fine-tuning was performed using QLoRA (Quantized LoRA), an extension of LoRA that introduces quantization for enhanced parameter efficiency.
|
23 |
+
The model benefits from 4-bit NormalFloat (NF4) quantization and Double Quantization techniques, ensuring optimized performance.
|
24 |
+
|
25 |
+
## Use Cases:
|
26 |
+
|
27 |
+
* **Automating Cover Letter Creation:** Llama2_7B_Cover_letter_generator can be used to rapidly generate cover letters for a wide range of job openings, saving time and effort for job seekers.
|
28 |
+
|
29 |
+
## Performance:
|
30 |
+
|
31 |
+
* Llama2_7B_Cover_letter_generator exhibits impressive performance in generating context-aware cover letters with high coherence and relevance to job descriptions.
|
32 |
+
* It maintains a low perplexity score, indicating its ability to generate text that aligns well with user input and desired contexts.
|
33 |
+
* The model's quantization techniques enhance its efficiency without significantly compromising performance.
|
34 |
+
|
35 |
+
## Limitations:
|
36 |
+
|
37 |
+
While the model excels in generating cover letters, it may occasionally produce text that requires minor post-processing for perfection.
|
38 |
+
+ It may not fully capture highly specific or niche job requirements, and some manual customization might be necessary for certain applications.
|
39 |
+
+ Llama2_7B_Cover_letter_generator's performance may vary depending on the complexity and uniqueness of the input prompts.
|
40 |
+
+ Users should be mindful of potential biases in the generated content and perform appropriate reviews to ensure inclusivity and fairness.
|
41 |
+
|
42 |
## Training procedure
|
43 |
|
44 |
|
|
|
56 |
|
57 |
|
58 |
- PEFT 0.4.0
|
59 |
+
|
60 |
+
|
61 |
+
## How to Get Started with the Model
|
62 |
+
|
63 |
+
```
|
64 |
+
! huggingface-cli login
|
65 |
+
```
|
66 |
+
|
67 |
+
```python
|
68 |
+
from transformers import pipeline
|
69 |
+
from transformers import AutoTokenizer
|
70 |
+
from peft import PeftModel, PeftConfig
|
71 |
+
from transformers import AutoModelForCausalLM , BitsAndBytesConfig
|
72 |
+
import torch
|
73 |
+
|
74 |
+
#config = PeftConfig.from_pretrained("ayoubkirouane/Llama2_13B_startup_hf")
|
75 |
+
bnb_config = BitsAndBytesConfig(
|
76 |
+
load_in_4bit=True,
|
77 |
+
bnb_4bit_quant_type="nf4",
|
78 |
+
bnb_4bit_compute_dtype=getattr(torch, "float16"),
|
79 |
+
bnb_4bit_use_double_quant=False)
|
80 |
+
model = AutoModelForCausalLM.from_pretrained(
|
81 |
+
"meta-llama/Llama-2-7b-hf",
|
82 |
+
quantization_config=bnb_config,
|
83 |
+
device_map={"": 0})
|
84 |
+
model.config.use_cache = False
|
85 |
+
model.config.pretraining_tp = 1
|
86 |
+
model = PeftModel.from_pretrained(model, "TuningAI/Llama2_7B_Cover_letter_generator")
|
87 |
+
tokenizer = AutoTokenizer.from_pretrained("meta-llama/Llama-2-7b-hf" , trust_remote_code=True)
|
88 |
+
tokenizer.pad_token = tokenizer.eos_token
|
89 |
+
tokenizer.padding_side = "right"
|
90 |
+
while 1:
|
91 |
+
input_text = input(">>>")
|
92 |
+
logging.set_verbosity(logging.CRITICAL)
|
93 |
+
prompt = f"### Instruction\n{system_message}.\n ###Input \n\n{input_text}. ### Output:"
|
94 |
+
pipe = pipeline(task="text-generation", model=model, tokenizer=tokenizer,max_length=512)
|
95 |
+
result = pipe(prompt)
|
96 |
+
print(result[0]['generated_text'].replace(prompt, ''))
|
97 |
+
```
|