Upload Lex Fridman chatbot model

README.md

@@ -0,0 +1,140 @@
+---
+language: en
+license: mit
+tags:
+- conversational
+- chatbot
+- lex-fridman
+- dialogue
+- pytorch
+- transformers
+- fine-tuned
+base_model: microsoft/DialoGPT-medium
+datasets:
+- custom
+widget:
+- text: "Human: What do you think about artificial intelligence?\n\nLex:"
+  example_title: "AI Discussion"
+- text: "Human: How do you approach difficult conversations?\n\nLex:"
+  example_title: "Conversation Approach"
+- text: "Human: What's your view on the future of technology?\n\nLex:"
+  example_title: "Future of Technology"
+---
+
+# Lex Fridman AI Chatbot
+
+This is a fine-tuned conversational AI model that mimics the conversational style of Lex Fridman, host of the Lex Fridman Podcast. The model was fine-tuned on podcast transcript data to capture Lex's thoughtful, philosophical, and engaging conversation style.
+
+## Model Details
+
+- **Base Model**: microsoft/DialoGPT-medium
+- **Model Type**: Conversational AI / Chatbot
+- **Language**: English
+- **Parameters**: ~117M
+- **Training Data**: Lex Fridman podcast transcripts
+- **Fine-tuning**: Custom dialogue dataset with Human-Lex conversation pairs
+
+## Training Details
+
+- **Training Examples**: 3,100
+- **Validation Examples**: 345
+- **Training Duration**: 0:20:30.474449
+- **Final Loss**: 4.1058
+- **Epochs**: 5
+- **Hardware**: NVIDIA A100 GPUs
+- **Framework**: PyTorch + Transformers
+
+## Usage
+
+```python
+from transformers import AutoTokenizer, AutoModelForCausalLM
+import torch
+
+# Load model and tokenizer
+tokenizer = AutoTokenizer.from_pretrained("your-username/lex-fridman-chatbot")
+model = AutoModelForCausalLM.from_pretrained("your-username/lex-fridman-chatbot")
+
+# Generate response
+def chat_with_lex(question):
+    prompt = f"Human: {question}\n\nLex:"
+    inputs = tokenizer.encode(prompt, return_tensors="pt")
+    
+    with torch.no_grad():
+        outputs = model.generate(
+            inputs,
+            max_length=inputs.shape[1] + 100,
+            temperature=0.8,
+            do_sample=True,
+            top_p=0.9,
+            pad_token_id=tokenizer.eos_token_id
+        )
+    
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return response.split("Lex:")[-1].strip()
+
+# Example usage
+response = chat_with_lex("What do you think about artificial intelligence?")
+print(response)
+```
+
+## Sample Conversations
+
+**Human**: What do you think about artificial intelligence?
+**Lex**: I think it's a big idea that we're seeing in the way computers are becoming self aware.
+
+**Human**: How do you approach difficult conversations?
+**Lex**: You approach them with your hands and you have a conversation with the viewer.
+
+## Web Interface
+
+This model comes with a beautiful Streamlit web interface for easy interaction:
+
+```bash
+streamlit run web_app/lex_chatbot_app.py
+```
+
+## Model Architecture
+
+- **Architecture**: GPT-2 based transformer (DialoGPT)
+- **Vocabulary Size**: 50,257 tokens
+- **Context Length**: 1024 tokens
+- **Precision**: FP16 for efficient inference
+- **Multi-GPU**: Supports distributed inference
+
+## Limitations and Biases
+
+- The model is trained on a limited dataset from podcast transcripts
+- May reflect biases present in the training data
+- Responses are generated based on patterns learned from transcripts
+- Not suitable for providing factual information or professional advice
+- Best used for casual conversation and entertainment
+
+## Ethical Considerations
+
+- This model is for educational and entertainment purposes
+- Does not represent the actual views or opinions of Lex Fridman
+- Should not be used to impersonate or misrepresent the real person
+- Users should be transparent about AI-generated content
+
+## Citation
+
+```bibtex
+@misc{lex-fridman-chatbot,
+  title={Lex Fridman AI Chatbot},
+  author={Your Name},
+  year={2025},
+  publisher={Hugging Face},
+  url={https://huggingface.co/your-username/lex-fridman-chatbot}
+}
+```
+
+## License
+
+MIT License - See LICENSE file for details.
+
+## Acknowledgments
+
+- Lex Fridman for the inspiring podcast content
+- Microsoft for the DialoGPT base model
+- Hugging Face for the transformers library
+- The open-source AI community

checkpoint-965/config.json

@@ -0,0 +1,38 @@
+{
+  "_name_or_path": "microsoft/DialoGPT-medium",
+  "activation_function": "gelu_new",
+  "architectures": [
+    "GPT2LMHeadModel"
+  ],
+  "attn_pdrop": 0.1,
+  "bos_token_id": 50256,
+  "embd_pdrop": 0.1,
+  "eos_token_id": 50256,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "model_type": "gpt2",
+  "n_ctx": 1024,
+  "n_embd": 1024,
+  "n_head": 16,
+  "n_inner": null,
+  "n_layer": 24,
+  "n_positions": 1024,
+  "reorder_and_upcast_attn": false,
+  "resid_pdrop": 0.1,
+  "scale_attn_by_inverse_layer_idx": false,
+  "scale_attn_weights": true,
+  "summary_activation": null,
+  "summary_first_dropout": 0.1,
+  "summary_proj_to_labels": true,
+  "summary_type": "cls_index",
+  "summary_use_proj": true,
+  "task_specific_params": {
+    "conversational": {
+      "max_length": 1000
+    }
+  },
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.46.2",
+  "use_cache": true,
+  "vocab_size": 50257
+}

checkpoint-965/generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 50256,
+  "eos_token_id": 50256,
+  "transformers_version": "4.46.2"
+}

checkpoint-965/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8913fc51f3d8b7ad1e272dbcc060fc84d4a5da6714b71920e47e79462291df90
+size 709676664

checkpoint-965/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4fea19ed2d0a009b3aeba1700053d91a99cf9e0a0bf6ce08c3db0de6fa4c266
+size 1419536634

checkpoint-965/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37b4712e69b70b808f19db8c1b2273bf2a7379d41a2b4a53635c9d2fcd761d5d
+size 14244

checkpoint-965/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1717724523643f30deded7cd205d277ea4e367e8f3b5dffcb921f61c48f2363a
+size 1064

checkpoint-965/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|endoftext|>",
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-965/tokenizer_config.json

@@ -0,0 +1,23 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "50256": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|endoftext|>",
+  "chat_template": "{% for message in messages %}{{ message.content }}{{ eos_token }}{% endfor %}",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "model_max_length": 1024,
+  "pad_token": "<|endoftext|>",
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>"
+}

checkpoint-965/trainer_state.json

@@ -0,0 +1,184 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.985806451612904,
+  "eval_steps": 500,
+  "global_step": 965,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 60.0,
+      "learning_rate": 1.5e-05,
+      "loss": 6.9241,
+      "step": 50
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 28.375,
+      "learning_rate": 3e-05,
+      "loss": 6.2276,
+      "step": 100
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 20.125,
+      "learning_rate": 2.8265895953757228e-05,
+      "loss": 4.8431,
+      "step": 150
+    },
+    {
+      "epoch": 1.033548387096774,
+      "grad_norm": 18.0,
+      "learning_rate": 2.653179190751445e-05,
+      "loss": 4.1853,
+      "step": 200
+    },
+    {
+      "epoch": 1.2916129032258064,
+      "grad_norm": 15.4375,
+      "learning_rate": 2.4797687861271675e-05,
+      "loss": 3.9644,
+      "step": 250
+    },
+    {
+      "epoch": 1.5496774193548388,
+      "grad_norm": 24.0,
+      "learning_rate": 2.3063583815028902e-05,
+      "loss": 3.838,
+      "step": 300
+    },
+    {
+      "epoch": 1.8077419354838709,
+      "grad_norm": 18.125,
+      "learning_rate": 2.132947976878613e-05,
+      "loss": 3.7908,
+      "step": 350
+    },
+    {
+      "epoch": 2.067096774193548,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.9595375722543353e-05,
+      "loss": 3.7595,
+      "step": 400
+    },
+    {
+      "epoch": 2.3251612903225807,
+      "grad_norm": 19.625,
+      "learning_rate": 1.7861271676300577e-05,
+      "loss": 3.7634,
+      "step": 450
+    },
+    {
+      "epoch": 2.5832258064516127,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.6127167630057804e-05,
+      "loss": 3.7074,
+      "step": 500
+    },
+    {
+      "epoch": 2.5832258064516127,
+      "eval_loss": 3.7095940113067627,
+      "eval_perplexity": 40.784299067771805,
+      "eval_runtime": 216.8944,
+      "eval_samples_per_second": 1.591,
+      "eval_steps_per_second": 0.798,
+      "step": 500
+    },
+    {
+      "epoch": 2.841290322580645,
+      "grad_norm": 14.875,
+      "learning_rate": 1.439306358381503e-05,
+      "loss": 3.6857,
+      "step": 550
+    },
+    {
+      "epoch": 3.1006451612903225,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.2658959537572255e-05,
+      "loss": 3.6973,
+      "step": 600
+    },
+    {
+      "epoch": 3.358709677419355,
+      "grad_norm": 23.25,
+      "learning_rate": 1.092485549132948e-05,
+      "loss": 3.7022,
+      "step": 650
+    },
+    {
+      "epoch": 3.616774193548387,
+      "grad_norm": 14.5,
+      "learning_rate": 9.190751445086706e-06,
+      "loss": 3.6774,
+      "step": 700
+    },
+    {
+      "epoch": 3.8748387096774195,
+      "grad_norm": 14.75,
+      "learning_rate": 7.45664739884393e-06,
+      "loss": 3.6879,
+      "step": 750
+    },
+    {
+      "epoch": 4.134193548387096,
+      "grad_norm": 15.625,
+      "learning_rate": 5.722543352601156e-06,
+      "loss": 3.6654,
+      "step": 800
+    },
+    {
+      "epoch": 4.392258064516129,
+      "grad_norm": 13.8125,
+      "learning_rate": 3.988439306358381e-06,
+      "loss": 3.6742,
+      "step": 850
+    },
+    {
+      "epoch": 4.650322580645161,
+      "grad_norm": 16.75,
+      "learning_rate": 2.254335260115607e-06,
+      "loss": 3.6861,
+      "step": 900
+    },
+    {
+      "epoch": 4.908387096774193,
+      "grad_norm": 17.5,
+      "learning_rate": 5.202312138728324e-07,
+      "loss": 3.6569,
+      "step": 950
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 965,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.867827744702464e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-965/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3849d00172e1d207cba3f4ac3120165d22f71ccdd4fed2cd97e66dfc3358dc2a
+size 5240

config.json

@@ -0,0 +1,38 @@
+{
+  "_name_or_path": "microsoft/DialoGPT-medium",
+  "activation_function": "gelu_new",
+  "architectures": [
+    "GPT2LMHeadModel"
+  ],
+  "attn_pdrop": 0.1,
+  "bos_token_id": 50256,
+  "embd_pdrop": 0.1,
+  "eos_token_id": 50256,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "model_type": "gpt2",
+  "n_ctx": 1024,
+  "n_embd": 1024,
+  "n_head": 16,
+  "n_inner": null,
+  "n_layer": 24,
+  "n_positions": 1024,
+  "reorder_and_upcast_attn": false,
+  "resid_pdrop": 0.1,
+  "scale_attn_by_inverse_layer_idx": false,
+  "scale_attn_weights": true,
+  "summary_activation": null,
+  "summary_first_dropout": 0.1,
+  "summary_proj_to_labels": true,
+  "summary_type": "cls_index",
+  "summary_use_proj": true,
+  "task_specific_params": {
+    "conversational": {
+      "max_length": 1000
+    }
+  },
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.46.2",
+  "use_cache": true,
+  "vocab_size": 50257
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 50256,
+  "eos_token_id": 50256,
+  "transformers_version": "4.46.2"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8913fc51f3d8b7ad1e272dbcc060fc84d4a5da6714b71920e47e79462291df90
+size 709676664

special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|endoftext|>",
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer_config.json

@@ -0,0 +1,23 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "50256": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|endoftext|>",
+  "chat_template": "{% for message in messages %}{{ message.content }}{{ eos_token }}{% endfor %}",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "model_max_length": 1024,
+  "pad_token": "<|endoftext|>",
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3849d00172e1d207cba3f4ac3120165d22f71ccdd4fed2cd97e66dfc3358dc2a
+size 5240

training_summary.json

@@ -0,0 +1,11 @@
+{
+  "model_name": "microsoft/DialoGPT-medium",
+  "training_loss": 4.105761517144238,
+  "training_duration": "0:20:30.474449",
+  "num_epochs": 5,
+  "num_gpus": 2,
+  "training_examples": 3100,
+  "validation_examples": 345,
+  "completed_at": "2025-06-20T04:32:48.471531",
+  "output_dir": "models/lex_chatbot"
+}