Upload 124M GPT with symbolic reasoning distillation

README.md

@@ -0,0 +1,20 @@
+---
+license: apache-2.0
+language: [en]
+tags: [text-generation, gpt2, knowledge-distillation, symbolic-reasoning, chain-of-thought, from-scratch]
+datasets: [HuggingFaceFW/fineweb-edu, openai/gsm8k]
+pipeline_tag: text-generation
+---
+# 124M GPT with Symbolic Reasoning Distillation
+
+Trained **from scratch** on mixed data with **dual-alpha distillation**:
+
+| Stream | Dataset | Alpha | Purpose |
+|--------|---------|-------|---------|
+| General | FineWeb-Edu | 0.2 | Language modeling, light teacher guidance |
+| **Reasoning** | **GSM8K chain-of-thought** | **0.8** | **Heavy distillation: teacher guides step-by-step math reasoning** |
+
+- **Teacher**: SmolLM-135M-Instruct (frozen)
+- **Time**: ~75 min on 1x A100
+- **Tokens**: 327,680,000 (0 reasoning / 20,000 general batches)
+- **Best loss**: 186.6474

chat_template.jinja

@@ -0,0 +1,4 @@
+{% for message in messages %}{{'<|im_start|>' + message['role'] + '
+' + message['content'] + '<|im_end|>' + '
+'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant
+' }}{% endif %}

config.json

@@ -0,0 +1,34 @@
+{
+  "activation_function": "gelu_new",
+  "add_cross_attention": false,
+  "architectures": [
+    "GPT2LMHeadModel"
+  ],
+  "attn_pdrop": 0.0,
+  "bos_token_id": 1,
+  "dtype": "float32",
+  "embd_pdrop": 0.0,
+  "eos_token_id": 2,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "model_type": "gpt2",
+  "n_embd": 768,
+  "n_head": 12,
+  "n_inner": null,
+  "n_layer": 12,
+  "n_positions": 512,
+  "pad_token_id": null,
+  "reorder_and_upcast_attn": false,
+  "resid_pdrop": 0.0,
+  "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,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.0.0",
+  "use_cache": true,
+  "vocab_size": 49152
+}

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "output_attentions": false,
+  "output_hidden_states": false,
+  "transformers_version": "5.0.0",
+  "use_cache": true
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80740b3aefcec895c40c43103c41873c4b57f8cf76dbe85beac6487400bcfb78
+size 492806784

raw_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5168c69e9f8a5df951d0e6900b3a30627680de7abe822a4ee2547d7c72029019
+size 492414052

tokenizer_config.json

@@ -0,0 +1,17 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": "<|im_start|>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>"
+  ],
+  "is_local": false,
+  "model_max_length": 1000000,
+  "pad_token": "<|im_end|>",
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": "<|endoftext|>",
+  "vocab_size": 49152
+}

training_config.json

@@ -0,0 +1,13 @@
+{
+  "teacher": "HuggingFaceTB/SmolLM-135M-Instruct",
+  "general_data": "HuggingFaceFW/fineweb-edu/sample-10BT",
+  "reasoning_data": "openai/gsm8k",
+  "alpha_general": 0.2,
+  "alpha_reasoning": 0.8,
+  "reasoning_ratio": 0.25,
+  "steps": 5000,
+  "tokens": 327680000,
+  "best_loss": 186.64740371704102,
+  "reasoning_batches": 0,
+  "general_batches": 20000
+}