{ "model": "Qwen/Qwen2.5-3B", "renderer": "role_colon", "batch_sha256": "27ee83599a2892f321ba2668fa343e3a4608fe2c30660732caceeed1820cd9b7", "ranks": { "1": { "create_seconds": 6.146354913711548, "info_after_create": { "type": "get_info", "model_data": { "arch": null, "model_name": "Qwen/Qwen2.5-3B", "tokenizer_id": null }, "model_id": "858fb34c3b2a340a600975ef5aaa34c2", "optimizer_config": { "type": "adamw" }, "is_lora": true, "lora_rank": 1, "model_name": "Qwen/Qwen2.5-3B" }, "initial_output_sha256": "5dd1397c851a36bcc9b3d4e3f0cb8007ca689d0a7805cd53dab3bd3fb5116332", "initial_logprobs_head": [ -3.346742630004883, -7.752749919891357, -1.9953076839447021, -9.7294282913208, -3.903634548187256 ], "initial_wnll": 0.4171471064994421, "init_state_path": "tinker://858fb34c3b2a340a600975ef5aaa34c2/weights/run_b89b93886b-init-r1", "adapter_sha256": null, "fb_loss_metrics": { "response_length:mean": 210.0, "loss:sum": 25.028827667236328, "loss:mean": 0.11918489365350632, "tokens:sum": 210.0, "n_sequences:sum": 1.0 }, "optim_step_response": { "metrics": { "update_successful:mean": 1.0, "timing/optimizer_step": 0.0, "timing/idle_wait_s": 17.548156912000195, "timing/idle_wait_calls": 9.0, "timing/prepare_batch_s": 0.00028734299996813206, "timing/prepare_batch_calls": 2.0, "timing/forward_only_s": 0.1579137090000131, "timing/forward_only_calls": 1.0, "timing/build_outputs_s": 0.00015779699992890528, "timing/build_outputs_calls": 2.0, "timing/save_checkpoint_s": 0.7497932929999251, "timing/save_checkpoint_calls": 1.0, "timing/persist_checkpoint_s": 10.209261041000104, "timing/persist_checkpoint_calls": 1.0, "timing/forward_backward_s": 0.4495170270000699, "timing/forward_backward_calls": 1.0, "timing/optim_step_s": 0.03182894099995792, "timing/optim_step_calls": 1.0, "timing/trainer_step_s": 0.4813459680000278, "grad_norm:mean": 539.6019897460938 } }, "info_after_update": { "type": "get_info", "model_data": { "arch": null, "model_name": "Qwen/Qwen2.5-3B", "tokenizer_id": null }, "model_id": "858fb34c3b2a340a600975ef5aaa34c2", "optimizer_config": { "type": "adamw" }, "is_lora": true, "lora_rank": 1, "model_name": "Qwen/Qwen2.5-3B" }, "post_update_wnll": 0.17814002547141475, "update_ok": true, "restore_seconds": 13.577327966690063, "info_after_restore": { "type": "get_info", "model_data": { "arch": null, "model_name": "Qwen/Qwen2.5-3B", "tokenizer_id": null }, "model_id": "42c77cdc-4ab6-473c-b967-a742aeabdd7c:train:0", "optimizer_config": { "type": "adamw" }, "is_lora": true, "lora_rank": 1, "model_name": "Qwen/Qwen2.5-3B" }, "restored_output_sha256": "5dd1397c851a36bcc9b3d4e3f0cb8007ca689d0a7805cd53dab3bd3fb5116332", "restored_max_abs_diff": 0.0 }, "32": { "create_seconds": 6.5747363567352295, "info_after_create": { "type": "get_info", "model_data": { "arch": null, "model_name": "Qwen/Qwen2.5-3B", "tokenizer_id": null }, "model_id": "879b6d3bac21f78eff5ba092b3b3b86b", "optimizer_config": { "type": "adamw" }, "is_lora": true, "lora_rank": 32, "model_name": "Qwen/Qwen2.5-3B" }, "initial_output_sha256": "5dd1397c851a36bcc9b3d4e3f0cb8007ca689d0a7805cd53dab3bd3fb5116332", "initial_logprobs_head": [ -3.346742630004883, -7.752749919891357, -1.9953076839447021, -9.7294282913208, -3.903634548187256 ], "initial_wnll": 0.4171471064994421, "init_state_path": "tinker://879b6d3bac21f78eff5ba092b3b3b86b/weights/run_b89b93886b-init-r32", "adapter_sha256": null, "fb_loss_metrics": { "response_length:mean": 210.0, "loss:sum": 25.028827667236328, "loss:mean": 0.11918489365350632, "tokens:sum": 210.0, "n_sequences:sum": 1.0 }, "optim_step_response": { "metrics": { "update_successful:mean": 1.0, "timing/optimizer_step": 0.0, "timing/idle_wait_s": 28.368350185999702, "timing/idle_wait_calls": 12.0, "timing/prepare_batch_s": 0.00046207400009734556, "timing/prepare_batch_calls": 3.0, "timing/forward_only_s": 0.2955712230000245, "timing/forward_only_calls": 2.0, "timing/build_outputs_s": 0.00023907100012365845, "timing/build_outputs_calls": 3.0, "timing/save_checkpoint_s": 1.2914502959999936, "timing/save_checkpoint_calls": 1.0, "timing/persist_checkpoint_s": 9.74062091099995, "timing/persist_checkpoint_calls": 1.0, "timing/forward_backward_s": 0.3840824760000032, "timing/forward_backward_calls": 1.0, "timing/optim_step_s": 0.02374846999998681, "timing/optim_step_calls": 1.0, "timing/trainer_step_s": 0.40783094599999004, "grad_norm:mean": 89.56376647949219 } }, "info_after_update": { "type": "get_info", "model_data": { "arch": null, "model_name": "Qwen/Qwen2.5-3B", "tokenizer_id": null }, "model_id": "879b6d3bac21f78eff5ba092b3b3b86b", "optimizer_config": { "type": "adamw" }, "is_lora": true, "lora_rank": 32, "model_name": "Qwen/Qwen2.5-3B" }, "post_update_wnll": 0.16179626008694564, "update_ok": true, "restore_seconds": 142.7230019569397, "info_after_restore": { "type": "get_info", "model_data": { "arch": null, "model_name": "Qwen/Qwen2.5-3B", "tokenizer_id": null }, "model_id": "67a59222-11b9-499b-b796-f728c2612317:train:0", "optimizer_config": { "type": "adamw" }, "is_lora": true, "lora_rank": 32, "model_name": "Qwen/Qwen2.5-3B" }, "restored_output_sha256": "5dd1397c851a36bcc9b3d4e3f0cb8007ca689d0a7805cd53dab3bd3fb5116332", "restored_max_abs_diff": 0.0 } }, "fresh_outputs_identical_across_ranks": true, "seconds": 217.0920534133911 }