{
  "schema_version": 3,
  "revision": 8,
  "revision_id": "r008",
  "release": "Pollock 1.6",
  "model_id": "SlayerLab/pollock-mini-lm-125m",
  "draft": false,
  "publication": {
    "weights_commit": "f8bd6fe8f95129f4f4c5d4177fae4ce35d222d04",
    "version_tag": "v1.6",
    "publication_commit_pending": false
  },
  "source_checkpoint": {
    "role": "best-full-validation",
    "sha256": "25b8cac7d5142c75e1be0b42fa24d980896e6f911e912c883566747a3baa56f0",
    "iteration": 32000,
    "tokens_seen": 15728561562,
    "native_unique_trainable_parameters": 127820544
  },
  "architecture": {
    "n_layer": 16,
    "n_head": 12,
    "n_embd": 768,
    "n_inner": 2144,
    "mlp_variant": "swiglu",
    "block_size": 2048,
    "vocab_size": 12288,
    "dropout": 0.0,
    "bias": false,
    "tied_word_embeddings": true
  },
  "training": {
    "dataset": "SlayerLab/minimal-en-corpus-5b",
    "dataset_revision": "38bebbd2ae36858c25aa0799fd3296fbfa9432aa",
    "training_tokens": 5396605407,
    "validation_tokens": 5238223,
    "tokenizer_sha256": "3733307577230bb4802d2c774d8e2323f7f64e4139c13712736a57daf91bdda1",
    "effective_batch_tokens": 491520,
    "optimizer": "Muon plus fused AdamW",
    "precision": "bfloat16",
    "hardware": "4x NVIDIA H100 80GB HBM3"
  },
  "evaluation": {
    "protocol": "complete validation split plus fixed corpus-wide shuffled train sample",
    "train_subset_tokens": 2457600,
    "validation_targets": 5238222,
    "selected_train_loss_rounded": 2.4182,
    "selected_validation_loss": 2.450774391943309,
    "benchmark_status": "completed",
    "benchmark_harness": "lm-evaluation-harness 0.4.12",
    "benchmark_num_fewshot": 0,
    "benchmark_batch_size": 8,
    "benchmark_checkpoint_sha256": "1c1002909119b4913841ee29efd44bd86d62d81f2ae27de01602b64d4af0dbec",
    "results_file": "benchmarks/english.json"
  },
  "fixed_inference": {
    "status": "completed",
    "directory": "inference-samples/r008",
    "device": "cpu",
    "dtype": "float32",
    "seed": 1337,
    "prompt_count": 4
  },
  "conversion": {
    "status": "completed",
    "target_class": "PollockForCausalLM",
    "transformers_version": "5.15.1",
    "checkpoint_sha256": "25b8cac7d5142c75e1be0b42fa24d980896e6f911e912c883566747a3baa56f0",
    "model_sha256": "1c1002909119b4913841ee29efd44bd86d62d81f2ae27de01602b64d4af0dbec",
    "unique_serialized_parameters": 127975936,
    "validation_probe_shape": [2, 64],
    "max_absolute_logit_error": 0.0
  },
  "artifacts": {
    "README.md": {"sha256": "6dc194e8f28bcdf4a3ff8890097a580bea300a45147808f58a97e56f2558caea"},
    "CHANGELOG.md": {"sha256": "f6fb59b4586c9c84f4a800865370c1889052ecba9bf6ff233e31e7ae2118211c"},
    "LICENSE.md": {"sha256": "eb699f26e2583809f58d84f2ca49779f2d39fe5fb455b54a78c0601d25cbcd62"},
    "config.json": {"sha256": "02372e3998549d69d8253275b468022711db3e262bc92ebc9338e5748c54be84"},
    "conversion_manifest.json": {"sha256": "a9a9a477d5c1a810e1f8c51e5cefd043a4ea90c5da381f452c81c8eabfb3ea0f"},
    "generation_config.json": {"sha256": "435beb27be51f0ed054f4a011e5109d125cdadc118b8799b18b155cc798d94d2"},
    "model.safetensors": {"sha256": "1c1002909119b4913841ee29efd44bd86d62d81f2ae27de01602b64d4af0dbec"},
    "modeling_pollock.py": {"sha256": "a078f5a8186dd52e4d128affce618c1a38b88e4ce52d000e5b5d9f5a73403171"},
    "tokenizer.json": {"sha256": "3733307577230bb4802d2c774d8e2323f7f64e4139c13712736a57daf91bdda1"},
    "tokenizer_config.json": {"sha256": "39a075b24f05384c6d759521563d879cab4bb80b0fe94381e6e792966ab4c2ba"},
    "training-history/r008.md": {"sha256": "9d3590ea89fae9fd39241b6eef307726f0b779af8637d0cd76684ca581518d2b"},
    "benchmarks/english.json": {"sha256": "f253da055603b7117df697a5c432301b68983bdb9f97b188ec84e0e795768f6d"},
    "inference-samples/r008/README.md": {"sha256": "f9f3ca69c57b20acd7fe635d4e25ca0ba22d3470038d165ad166f36465f7fe7f"},
    "inference-samples/r008/results.json": {"sha256": "80c3096451ce68dbb0f50e99123d1aedf826a5669da09a0bbb455e4a07fe5750"},
    "inference-samples/r008/generate.py": {"sha256": "79fc02c872587435e9f4cd88c89df7ef2ad9d6b2b1aabae68d3d09bc9afeeea3"},
    "inference-samples/r008/requirements.txt": {"sha256": "533e700863c1970e2706ad316b7fa1e7194374c484e75cc743a71e4ff624390a"}
  }
}