Upload a merged model in Huggingface

upload_merge.py

import argparse
import yaml
from huggingface_hub import ModelCard, HfApi
from jinja2 import Template

# Argument parser setup
parser = argparse.ArgumentParser(description='Create a repo on Huggingface for the model and upload it.')
parser.add_argument('model_name', type=str, help='Model name')
parser.add_argument('--new', action='store_true', default=True, help='Create new repo')
parser.add_argument('-y', '--yaml_path', type=str, required=True, help='Path to YAML configuration file')
parser.add_argument('-u', '--username', type=str, default="Technoculture", help='Hugging Face username')
parser.add_argument('-t', '--token', type=str, required=True, help='Hugging Face API token')
parser.add_argument('-m', '--is_moe', type=bool, default=False, help='Hugging Face API token')
args = parser.parse_args()

# Read YAML configuration file
with open(args.yaml_path, 'r') as file:
    yaml_config = file.read()
data = yaml.safe_load(yaml_config)
print(data)

# Branch selection and template setup
is_moe = args.is_moe
if not is_moe:
    template_text = """
---
license: apache-2.0
tags:
- merge
- mergekit
{%- for model in models %}
- {{ model }}
{%- endfor %}
---

# {{ model_name }}

{{ model_name }} is a merge of the following models:

{%- for model in models %}
* [{{ model }}](https://huggingface.co/{{ model }})
{%- endfor %}

## 🧩 Configuration

```yaml
{{- yaml_config -}}
```

## 💻 Usage

```python
!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "{{ username }}/{{ model_name }}"
messages = [{"role": "user", "content": "I am feeling sleepy these days"}]

tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    torch_dtype=torch.float16,
    device_map="auto",
)

outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
```
"""
        # Create a Jinja template object
    jinja_template = Template(template_text.strip())

    # Get list of models from config
    data = yaml.safe_load(yaml_config)
    if "models" in data:
        models = [data["models"][i]["model"] for i in range(len(data["models"])) if "parameters" in data["models"][i]]
    elif "parameters" in data:
        models = [data["slices"][0]["sources"][i]["model"] for i in range(len(data["slices"][0]["sources"]))]
    elif "slices" in data:
        models = [data["slices"][i]["sources"][0]["model"] for i in range(len(data["slices"]))]
    else:
        raise Exception("No models or slices found in yaml config")

    # Fill the template
    content = jinja_template.render(
        model_name=args.model_name,
        models=models,
        yaml_config=yaml_config,
        username=args.username,
    )
else:
    template_text = """
---
license: apache-2.0
tags:
- moe
- merge
{%- for model in models %}
- {{ model }}
{%- endfor %}
---

# {{ model_name }}

{{ model_name }} is a Mixure of Experts (MoE) made with the following models:

{%- for model in models %}
* [{{ model }}](https://huggingface.co/{{ model }})
{%- endfor %}

## 🧩 Configuration

```yaml
{{- yaml_config -}}
```

## 💻 Usage

```python
!pip install -qU transformers bitsandbytes accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "{{ username }}/{{ model_name }}"

tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    model_kwargs={"torch_dtype": torch.float16, "load_in_4bit": True},
)

messages = [{"role": "user", "content": "Explain what a Mixture of Experts is in less than 100 words."}]
prompt = pipeline.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
```
"""
    # Create a Jinja template object
    jinja_template = Template(template_text.strip())

    # Fill the template
    models = [model['source_model'] for model in data['experts']]

    content = jinja_template.render(
        model_name=args.model_name,
        models=models,
        yaml_config=yaml_config,
        username=args.username,
    )

# Save the model card
card = ModelCard(content)
card.save('merge/README.md')

# Defined in the secrets tab in Google Colab
api = HfApi(token=args.token)

if args.new:
    api.create_repo(
        repo_id=f"{args.username}/{args.model_name}",
        repo_type="model"
    )

# Upload merge folder
api.upload_folder(
    repo_id=f"{args.username}/{args.model_name}",
    folder_path="merge",
)