Instructions to use pzarzycki/hrm-text-1b-code-tools-sft with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Keras
How to use pzarzycki/hrm-text-1b-code-tools-sft with Keras:
# Available backend options are: "jax", "torch", "tensorflow". import os os.environ["KERAS_BACKEND"] = "jax" import keras model = keras.saving.load_model("hf://pzarzycki/hrm-text-1b-code-tools-sft") - KerasHub
How to use pzarzycki/hrm-text-1b-code-tools-sft with KerasHub:
import keras_hub # Create a Backbone model unspecialized for any task backbone = keras_hub.models.Backbone.from_preset("hf://pzarzycki/hrm-text-1b-code-tools-sft") - Keras
How to use pzarzycki/hrm-text-1b-code-tools-sft with Keras:
# Available backend options are: "jax", "torch", "tensorflow". import os os.environ["KERAS_BACKEND"] = "jax" import keras model = keras.saving.load_model("hf://pzarzycki/hrm-text-1b-code-tools-sft") - Notebooks
- Google Colab
- Kaggle
| { | |
| "config": { | |
| "data": { | |
| "context_length": 4096, | |
| "manifest": "/workspace/sft/stage-a/manifest.json", | |
| "response_only_loss": true, | |
| "tokenizer_json": "/workspace/models/hrm-text-1b/tokenizer.reference.json" | |
| }, | |
| "framework": "keras", | |
| "model": { | |
| "family": "hrm-text", | |
| "preset": "/workspace/models/hrm-text-1b", | |
| "rematerialization": true, | |
| "source_lock": "keras-hub.lock" | |
| }, | |
| "run_name": "hrm-stage-a-pilot", | |
| "training": { | |
| "backup_save_freq": 500, | |
| "beta_1": 0.9, | |
| "beta_2": 0.95, | |
| "checkpoint_save_freq": 20000, | |
| "clipnorm": 1.0, | |
| "cosine_min_lr_ratio": 0.1, | |
| "ema_momentum": 0.999, | |
| "gradient_accumulation_steps": 8, | |
| "learning_rate": 3e-05, | |
| "micro_batch_size": 1, | |
| "observability": { | |
| "every_n_batches": 10, | |
| "tensorboard_every_n_batches": 10 | |
| }, | |
| "output_dir": "artifacts/hrm-stage-a-pilot-seed17", | |
| "precision": "bfloat16", | |
| "seed": 17, | |
| "target_response_tokens": 10000000, | |
| "use_ema": true, | |
| "validation_steps": 128, | |
| "warmup_ratio": 0.03, | |
| "weight_decay": 0.1 | |
| } | |
| }, | |
| "config_sha256": "3e535c768bd399c8dbdbbb54d13ad03e1fc56d83bf6171cf3ed44a090f80327c", | |
| "manifest": { | |
| "active_token_ids": { | |
| "<|box_end|>": 11, | |
| "<|endoftext|>": 5, | |
| "<|im_end|>": 7, | |
| "<|im_start|>": 6, | |
| "<|object_ref_end|>": 9, | |
| "<|object_ref_start|>": 8, | |
| "<|quad_end|>": 13, | |
| "<|quad_start|>": 12 | |
| }, | |
| "condition_distribution": { | |
| "direct": 1157064 | |
| }, | |
| "data_schema_version": 2, | |
| "declared_license": "CC-BY-4.0", | |
| "decontamination": "complete", | |
| "decontamination_report": "decontamination.json", | |
| "decontamination_report_sha256": "3c4ace8d0ca0a668db2ebcea4ef6874499b6aefbb359d324dbfc694ac48f35ba", | |
| "deduplication": { | |
| "exact": "complete" | |
| }, | |
| "format": "hrm-code-tools-sft-v2", | |
| "migration": { | |
| "input_manifest_sha256": "18c43a551e6c62adf600deb9614d3057ebb1db6dd1ecff5e76b76f26ee375881", | |
| "input_train_sha256": "5d5b4a8f0fd37d4efb43db433c5d715f0f7ded228a07406e294e8cfed63f752a", | |
| "input_validation_sha256": "3101880dc4b750499617f24312984c05d23081ae68e37739366f9ffebf0cd0d0", | |
| "method": "v1 curated rows -> canonical v2 fields; re-split, deduplicate, and re-cap" | |
| }, | |
| "records": { | |
| "accepted": 1157064, | |
| "by_source": { | |
| "nvidia/OpenCodeInstruct": 1157064 | |
| }, | |
| "input_rows": 1157064, | |
| "rejected_context_limit": 0, | |
| "rejected_duplicate_id": 0, | |
| "rejected_exact_duplicate": 0, | |
| "response_tokens": 300785226, | |
| "serialized_tokens": 590957065, | |
| "train": 1133817, | |
| "validation": 23247 | |
| }, | |
| "selection": { | |
| "input_was_full_stage": true, | |
| "max_serialized_sequence_length": 4096 | |
| }, | |
| "serializer_version": "hrm-prefixlm-sft-v1", | |
| "sources": { | |
| "nvidia/OpenCodeInstruct": { | |
| "license": "CC-BY-4.0", | |
| "revision": "8f3ba5bafe4d6e8db46082cf7ae6741bc370604d" | |
| } | |
| }, | |
| "split": { | |
| "method": "sha256(source-id)%100", | |
| "validation_percent": 2 | |
| }, | |
| "stage": "stage-a", | |
| "tokenizer_json": "/data1/contributions/keras/hrm-text-1b-reference/tokenizer.json", | |
| "tokenizer_revision": "d08ed34cc79b3c788d7b5d81399511439a0e39af", | |
| "tokenizer_sha256": "ba520107dc9b27d25643e8f43fc1d9921038c86d3ada9dcaabf18526c5836d50", | |
| "train_jsonl": "train.jsonl", | |
| "train_sha256": "b6e340ea64568570b24e2d5f08c506a6f3221482d9fca6778a49c5577832171c", | |
| "transform_version": "hrm-sft-v2", | |
| "validation_jsonl": "validation.jsonl", | |
| "validation_sha256": "f8c52067e80f5a7edb8be22a08cbc20de421ae6f368878f9bab64abe88a1bcc4" | |
| }, | |
| "status": "preflight-passed" | |
| } | |