Upload initialized from-scratch smoke bootstrap artifacts (tokenizer + random-init checkpoint)

scratch/bootstrap_smoke_2026-03-23/checkpoints/config.json

@@ -0,0 +1,35 @@
+{
+  "activation_function": "gelu_new",
+  "add_cross_attention": false,
+  "architectures": [
+    "GPT2LMHeadModel"
+  ],
+  "attn_pdrop": 0.1,
+  "bos_token_id": 2,
+  "dtype": "float32",
+  "embd_pdrop": 0.1,
+  "eos_token_id": 3,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "model_type": "gpt2",
+  "n_ctx": 1024,
+  "n_embd": 256,
+  "n_head": 4,
+  "n_inner": null,
+  "n_layer": 4,
+  "n_positions": 1024,
+  "pad_token_id": 0,
+  "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,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.2.0",
+  "use_cache": false,
+  "vocab_size": 12000
+}

scratch/bootstrap_smoke_2026-03-23/checkpoints/generation_config.json

@@ -0,0 +1,10 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 2,
+  "eos_token_id": 3,
+  "output_attentions": false,
+  "output_hidden_states": false,
+  "pad_token_id": 0,
+  "transformers_version": "5.2.0",
+  "use_cache": true
+}

scratch/bootstrap_smoke_2026-03-23/checkpoints/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bc28abb0fd0db578b527ed1858fb19521b290d27080ae284bb44fa527fd23c0
+size 25979888

scratch/bootstrap_smoke_2026-03-23/checkpoints/resolved_training_config.json

@@ -0,0 +1,73 @@
+{
+  "global": {
+    "output_root": "model_development/runs/math-conjecture-scratch-smoke",
+    "seed": 17
+  },
+  "tokenizer": {
+    "tokenizer_dir": "model_development/runs/math-conjecture-scratch-smoke/tokenizer",
+    "vocab_size": 12000,
+    "min_frequency": 2,
+    "max_train_rows": 8000,
+    "special_tokens": [
+      "<pad>",
+      "<unk>",
+      "<s>",
+      "</s>",
+      "<|system|>",
+      "<|user|>",
+      "<|assistant|>"
+    ]
+  },
+  "model": {
+    "n_layer": 4,
+    "n_head": 4,
+    "n_embd": 256,
+    "n_positions": 1024,
+    "use_bf16": false,
+    "resid_pdrop": 0.1,
+    "embd_pdrop": 0.1,
+    "attn_pdrop": 0.1,
+    "initializer_range": 0.02
+  },
+  "data": {
+    "train_file": "data/releases/v1/train.parquet",
+    "validation_file": "data/releases/v1/validation.parquet",
+    "prompt_field": "prompt",
+    "target_field": "target",
+    "final_answer_field": "final_answer",
+    "proof_field": "proof_formal",
+    "max_seq_length": 1024,
+    "max_train_samples": 5000,
+    "max_eval_samples": 500,
+    "system_prompt": "You are NorthernTribe Research's math-conjecture solver.\n"
+  },
+  "training": {
+    "output_dir": "model_development/runs/math-conjecture-scratch-smoke/checkpoints",
+    "num_train_epochs": 1,
+    "max_steps": 20,
+    "per_device_train_batch_size": 1,
+    "per_device_eval_batch_size": 1,
+    "gradient_accumulation_steps": 4,
+    "learning_rate": 0.0003,
+    "weight_decay": 0.05,
+    "warmup_ratio": 0.02,
+    "lr_scheduler_type": "cosine",
+    "max_grad_norm": 1.0,
+    "gradient_checkpointing": false,
+    "logging_steps": 5,
+    "save_steps": 10,
+    "eval_steps": 10,
+    "save_total_limit": 2,
+    "dataloader_num_workers": 0,
+    "seed": 17
+  },
+  "hub": {
+    "push_to_hub": false,
+    "repo_id": "NorthernTribe-Research/math-conjecture-model",
+    "private": false,
+    "commit_message": "Upload scratch-smoke math-conjecture solver artifacts."
+  },
+  "credentials": {
+    "path": "huggingface-api-key.json"
+  }
+}

scratch/bootstrap_smoke_2026-03-23/checkpoints/tokenizer_config.json

@@ -0,0 +1,14 @@
+{
+  "backend": "tokenizers",
+  "bos_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": [
+    "<|system|>",
+    "<|user|>",
+    "<|assistant|>"
+  ],
+  "model_max_length": 1024,
+  "pad_token": "<pad>",
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": "<unk>"
+}

scratch/bootstrap_smoke_2026-03-23/scratch_init_summary.json

@@ -0,0 +1,13 @@
+{
+  "mode": "init_only",
+  "output_dir": "model_development/runs/math-conjecture-scratch-smoke/checkpoints",
+  "tokenizer_dir": "/home/ntr/Documents/huggingface/datasets/maths-conjuncture-solutions/model_development/runs/math-conjecture-scratch-smoke/tokenizer",
+  "rows_train": 5000,
+  "rows_validation": 500,
+  "max_seq_length": 1024,
+  "model": {
+    "total_parameters": 6493696,
+    "trainable_parameters": 6493696
+  },
+  "config_path": "model_development/configs/math_conjecture_scratch_smoke.yaml"
+}

scratch/bootstrap_smoke_2026-03-23/tokenizer/tokenizer_config.json

@@ -0,0 +1,14 @@
+{
+  "backend": "tokenizers",
+  "bos_token": "<s>",
+  "eos_token": "</s>",
+  "extra_special_tokens": [
+    "<|system|>",
+    "<|user|>",
+    "<|assistant|>"
+  ],
+  "model_max_length": 1024,
+  "pad_token": "<pad>",
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": "<unk>"
+}