Initial Upload

LICENSE.md

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

README.md

@@ -1,14 +1,32 @@
----
-title: Woq Inference
-emoji: 🐢
-colorFrom: purple
-colorTo: purple
-sdk: gradio
-sdk_version: 5.28.0
-app_file: app.py
-pinned: false
-license: apache-2.0
-short_description: A space where you can test WoQ quantized models using INC
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# PyTorch Weights-only-Quantization (WoQ)
+
+Inference scripts for pytorch weights-only-quantization
+
+## TEQ: a trainable equivalent transformation that preserves the FP32 precision in weight-only quantization
+
+### Install
+
+```
+conda create -n teq-inference python=3.11
+
+conda activate teq-inference
+
+conda install -c conda-forge gcc
+
+pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu
+
+pip install -r requirements.txt
+```
+
+### Usage
+
+```
+python teq_inference.py --base <base_model> --model_dir <path-to-woq-TEQ-quantized-model> --weights_file quantized_weight.pt --config_file qconfig.json --prompt "Tell me a joke" --device cpu
+```
+
+For example:
+
+```
+python teq_inference.py --base meta-llama/Llama-3.2-1B --model_dir ./meta-llama_Llama-3.2-1B-TEQ-int4-gs128-asym --weights_file quantized_weight.pt --config_file qconfig.json --prompt "Tell me a joke" --device cpu
+```
+

app.py

@@ -0,0 +1,77 @@
+import os
+import gradio as gr
+from huggingface_hub import list_repo_files, hf_hub_download
+import subprocess
+
+# Constants
+HF_USER = "fbaldassarri"
+TEQ_KEYWORD = "TEQ"
+
+def list_teq_models():
+    # List all repos with TEQ in their name
+    from huggingface_hub import list_repos
+    repos = list_repos(HF_USER)
+    return [repo.id for repo in repos if TEQ_KEYWORD in repo.id]
+
+def list_model_files(model_id):
+    # List files in the repo that are likely to be weights/config
+    files = list_repo_files(model_id)
+    weights = [f for f in files if f.endswith('.pt')]
+    configs = [f for f in files if f.endswith('.json')]
+    return weights, configs
+
+def run_teq_inference(model_id, weights_file, config_file, base_model, prompt, max_new_tokens, debug):
+    # Download files if not present
+    local_model_dir = f"./models/{model_id.replace('/', '_')}"
+    os.makedirs(local_model_dir, exist_ok=True)
+    weights_path = hf_hub_download(model_id, weights_file, local_dir=local_model_dir)
+    config_path = hf_hub_download(model_id, config_file, local_dir=local_model_dir)
+    # Call teq_inference.py as a subprocess for isolation
+    cmd = [
+        "python", "teq_inference.py",
+        "--model_dir", local_model_dir,
+        "--weights_file", weights_file,
+        "--config_file", config_file,
+        "--base_model", base_model,
+        "--prompt", prompt,
+        "--max_new_tokens", str(max_new_tokens),
+        "--device", "cpu"
+    ]
+    if debug:
+        cmd.append("--debug")
+    result = subprocess.run(cmd, capture_output=True, text=True)
+    # Extract generated text from logs
+    output = result.stdout + "\n" + result.stderr
+    # Try to find the generated text in logs
+    marker = "Generated text:"
+    if marker in output:
+        return output.split(marker)[-1].strip()
+    return output
+
+# Gradio UI
+def ui():
+    teq_models = list_teq_models()
+    with gr.Blocks() as demo:
+        gr.Markdown("# TEQ Quantized Model Inference Demo")
+        model_id = gr.Dropdown(teq_models, label="Select TEQ Model")
+        weights_file = gr.Textbox(label="Weights File (.pt)")
+        config_file = gr.Textbox(label="Config File (.json)")
+        base_model = gr.Textbox(label="Base Model Name", value="facebook/opt-350m")
+        prompt = gr.Textbox(label="Prompt", value="Once upon a time, a little girl")
+        max_new_tokens = gr.Slider(10, 512, value=100, label="Max New Tokens")
+        debug = gr.Checkbox(label="Debug Mode")
+        output = gr.Textbox(label="Generated Text", lines=10)
+        def update_files(model_id):
+            weights, configs = list_model_files(model_id)
+            return gr.update(choices=weights), gr.update(choices=configs)
+        model_id.change(update_files, inputs=model_id, outputs=[weights_file, config_file])
+        run_btn = gr.Button("Run Inference")
+        run_btn.click(
+            run_teq_inference,
+            inputs=[model_id, weights_file, config_file, base_model, prompt, max_new_tokens, debug],
+            outputs=output
+        )
+    return demo
+
+if __name__ == "__main__":
+    ui().launch()

requirements.txt

@@ -0,0 +1,9 @@
+accelerate==1.6.0
+datasets==3.5.0
+pydantic==2.11.3
+transformers==4.51.3
+neural_compressor==3.3.1
+intel-extension-for-pytorch==2.7.0
+gradio
+huggingface_hub
+

teq_inference.py

@@ -0,0 +1,162 @@
+import os
+import json
+import torch
+import argparse
+import logging
+from transformers import AutoTokenizer, AutoModelForCausalLM
+
+# Set up logging
+logging.basicConfig(level=logging.INFO, format='%(asctime)s [%(levelname)s] %(message)s')
+logger = logging.getLogger(__name__)
+
+def debug_weights_structure(weights_path):
+    """Examine the structure of the weights file to help debug loading issues"""
+    weights = torch.load(weights_path, map_location="cpu")
+    logger.info(f"Type of loaded weights: {type(weights)}")
+    if isinstance(weights, dict):
+        logger.info(f"Top-level keys: {list(weights.keys())}")
+        # Print a few sample keys to understand the structure
+        sample_keys = list(weights.keys())[:5]
+        for key in sample_keys:
+            logger.info(f"Sample key structure: {key} -> {type(weights[key])}")
+    return weights
+
+def main():
+    parser = argparse.ArgumentParser(description="Run inference with a TEQ-quantized model")
+    parser.add_argument("--model_dir", type=str, default=".",
+                        help="Directory containing quantized model files")
+    parser.add_argument("--weights_file", type=str, default="quantized_weight.pt",
+                        help="Name of the quantized weights file")
+    parser.add_argument("--config_file", type=str, default="qconfig.json",
+                        help="Name of the quantization config file")
+    parser.add_argument("--base_model", type=str, required=True,
+                        help="Original model name or path (for tokenizer and model architecture)")
+    parser.add_argument("--prompt", type=str, default="Once upon a time, a little girl",
+                        help="Text prompt for inference")
+    parser.add_argument("--max_new_tokens", type=int, default=100,
+                        help="Maximum number of new tokens to generate")
+    parser.add_argument("--device", type=str, default="cpu", choices=["cpu", "cuda", "xpu"],
+                        help="Device to run inference on")
+    parser.add_argument("--output_file", type=str, default=None,
+                        help="File to save the generated text to (optional)")
+    parser.add_argument("--debug", action="store_true",
+                        help="Print additional debug information")
+    args = parser.parse_args()
+
+    # Set up paths
+    weights_path = os.path.join(args.model_dir, args.weights_file)
+    config_path = os.path.join(args.model_dir, args.config_file)
+    
+    # Check if files exist
+    if not os.path.exists(weights_path):
+        raise FileNotFoundError(f"Quantized weights file not found: {weights_path}")
+    if not os.path.exists(config_path):
+        raise FileNotFoundError(f"Quantization config file not found: {config_path}")
+    
+    # Load tokenizer
+    logger.info(f"Loading tokenizer from {args.base_model}...")
+    tokenizer = AutoTokenizer.from_pretrained(args.base_model, trust_remote_code=True)
+    
+    # Examine the structure of the weights file
+    logger.info(f"Analyzing weights structure from {weights_path}...")
+    weights = debug_weights_structure(weights_path)
+    
+    # Load the base model directly (bypassing TEQ quantization)
+    logger.info(f"Loading base model from {args.base_model}...")
+    model = AutoModelForCausalLM.from_pretrained(args.base_model, trust_remote_code=True)
+    
+    # Print model's state_dict keys for debugging
+    if args.debug:
+        model_keys = list(model.state_dict().keys())
+        logger.info(f"Model has {len(model_keys)} keys in state_dict")
+        logger.info(f"Sample model keys: {model_keys[:5]}")
+    
+    # Check if weights contains 'state_dict' key and adjust accordingly
+    if 'state_dict' in weights:
+        logger.info("Found 'state_dict' key in weights file, extracting it...")
+        weights = weights['state_dict']
+    
+    # Try to match the weights to the model structure
+    try:
+        # First attempt: Direct loading
+        logger.info("Attempting to load weights directly...")
+        missing_keys, unexpected_keys = model.load_state_dict(weights, strict=False)
+        
+        if missing_keys:
+            logger.warning(f"Missing {len(missing_keys)} keys in state_dict")
+            if args.debug:
+                logger.warning(f"Sample missing keys: {missing_keys[:5]}")
+        
+        if unexpected_keys:
+            logger.warning(f"Found {len(unexpected_keys)} unexpected keys in state_dict")
+            if args.debug:
+                logger.warning(f"Sample unexpected keys: {unexpected_keys[:5]}")
+                
+        # Validate if we have critical missing keys
+        if len(missing_keys) > len(model.state_dict()) * 0.5:
+            logger.error("Too many missing keys! Weight loading may have failed")
+            
+    except Exception as e:
+        logger.error(f"Error loading weights: {str(e)}")
+        logger.info("Attempting to transform keys to match model structure...")
+        
+        # Create a transformed state_dict
+        transformed_weights = {}
+        
+        # Try removing 'module.' prefix
+        for key in weights:
+            if key.startswith('module.'):
+                transformed_weights[key[7:]] = weights[key]
+            else:
+                transformed_weights[key] = weights[key]
+        
+        # Try loading the transformed weights
+        missing_keys, unexpected_keys = model.load_state_dict(transformed_weights, strict=False)
+        logger.info(f"After transformation: {len(missing_keys)} missing keys, {len(unexpected_keys)} unexpected keys")
+    
+    # Put model in evaluation mode
+    model.eval()
+    
+    # Move model to specified device
+    device = args.device
+    logger.info(f"Moving model to {device}...")
+    model = model.to(device)
+    
+    # Optimize with IPEX if using Intel hardware
+    if device == "xpu":
+        try:
+            import intel_extension_for_pytorch as ipex
+            logger.info("Optimizing model with IPEX...")
+            model = ipex.optimize(model, dtype=torch.float16)
+        except ImportError:
+            logger.warning("IPEX not available, skipping optimization")
+    
+    # Run inference
+    logger.info(f"Generating text for prompt: '{args.prompt}'")
+    inputs = tokenizer(args.prompt, return_tensors="pt").to(device)
+    
+    # Generate text
+    with torch.no_grad():
+        output_ids = model.generate(
+            inputs["input_ids"],
+            max_new_tokens=args.max_new_tokens,
+            do_sample=True,
+            temperature=0.7,
+            top_p=0.9,
+        )
+    
+    # Decode the generated tokens
+    generated_text = tokenizer.batch_decode(output_ids, skip_special_tokens=True)[0]
+    logger.info("\nGenerated text:")
+    logger.info("-" * 50)
+    logger.info(generated_text)
+    logger.info("-" * 50)
+    
+    # Save to file if specified
+    if args.output_file:
+        with open(args.output_file, 'w') as f:
+            f.write(generated_text)
+        logger.info(f"Generated text saved to {args.output_file}")
+
+if __name__ == "__main__":
+    main()