{ "checkpoint_count": 7, "model_family": "qwen", "namespace": "cache_mask_v1_reviewer_subset_v1", "numpy_version": "2.3.5", "records": [ { "checkpoint_path": "runs/qwen-capacity-fixed-linear-t8-s1234/checkpoint/step_025000.pt", "checkpoint_sha256": "c1282c23b6e1edfd94ca99dd37560b8ebd4b9d6eb42ab1d1afe7e18b078bd58a", "checkpoint_size_bytes": 24170898, "ema_payload_present": false, "hub_slug": "qwen-capacity-fixed-linear-t8-s1234", "model_state_prefixes": [ "feedback_proj", "init_token", "latent_pooler", "mode_embed", "out_ln", "token0_head", "transform" ], "optimizer_payload_present": false, "private_string_count": 0, "status": "PASS", "top_level_keys": [ "ckpt_mode", "ema_shadow", "model_slim", "optim", "rng_state", "scaler", "scheduler", "step" ], "unsafe_globals": [ "numpy._core.multiarray._reconstruct", "numpy.dtype", "numpy.ndarray" ], "victim_weight_keys": 0 }, { "checkpoint_path": "runs/qwen-capacity-fixed-residual-mlp-t8-s1234/checkpoint/step_022000.pt", "checkpoint_sha256": "aaedd17ec6d1d11f3b384d1b7c396f195ed8170f41d0be0f9cfbcbbf97e10d3b", "checkpoint_size_bytes": 98346694, "ema_payload_present": false, "hub_slug": "qwen-capacity-fixed-residual-mlp-t8-s1234", "model_state_prefixes": [ "feedback_proj", "init_token", "latent_pooler", "mode_embed", "out_ln", "token0_head", "transform" ], "optimizer_payload_present": false, "private_string_count": 0, "status": "PASS", "top_level_keys": [ "ckpt_mode", "ema_shadow", "model_slim", "optim", "rng_state", "scaler", "scheduler", "step" ], "unsafe_globals": [ "numpy._core.multiarray._reconstruct", "numpy.dtype", "numpy.ndarray" ], "victim_weight_keys": 0 }, { "checkpoint_path": "runs/qwen-length-full-t32-s1234/checkpoint/step_025000.pt", "checkpoint_sha256": "707aae44d9c0fc32999b73fb8d7cae4d2ec4363bda8be05092ee483a0568b8e7", "checkpoint_size_bytes": 98354880, "ema_payload_present": false, "hub_slug": "qwen-length-full-t32-s1234", "model_state_prefixes": [ "feedback_proj", "init_token", "latent_pooler", "mode_embed", "out_ln", "token0_head", "transform" ], "optimizer_payload_present": false, "private_string_count": 0, "status": "PASS", "top_level_keys": [ "ckpt_mode", "ema_shadow", "model_slim", "optim", "rng_state", "scaler", "scheduler", "step" ], "unsafe_globals": [ "numpy._core.multiarray._reconstruct", "numpy.dtype", "numpy.ndarray" ], "victim_weight_keys": 0 }, { "checkpoint_path": "runs/qwen-prefix-direct-only-t8-s1234/checkpoint/step_025000.pt", "checkpoint_sha256": "bdc45e141c63825d7cf71d1649f5f5cae903855a219c5072a0f690471dc9dc0f", "checkpoint_size_bytes": 98354880, "ema_payload_present": false, "hub_slug": "qwen-prefix-direct-only-t8-s1234", "model_state_prefixes": [ "feedback_proj", "init_token", "latent_pooler", "mode_embed", "out_ln", "token0_head", "transform" ], "optimizer_payload_present": false, "private_string_count": 0, "status": "PASS", "top_level_keys": [ "ckpt_mode", "ema_shadow", "model_slim", "optim", "rng_state", "scaler", "scheduler", "step" ], "unsafe_globals": [ "numpy._core.multiarray._reconstruct", "numpy.dtype", "numpy.ndarray" ], "victim_weight_keys": 0 }, { "checkpoint_path": "runs/qwen-primary-final-t8-s1234/checkpoint/step_005000.pt", "checkpoint_sha256": "2053bbe8d644d118321c0aab0e702e2d9c04ae628eca67b476c6ec29162b693f", "checkpoint_size_bytes": 98354816, "ema_payload_present": false, "hub_slug": "qwen-primary-final-t8-s1234", "model_state_prefixes": [ "feedback_proj", "init_token", "latent_pooler", "mode_embed", "out_ln", "token0_head", "transform" ], "optimizer_payload_present": false, "private_string_count": 0, "status": "PASS", "top_level_keys": [ "ckpt_mode", "ema_shadow", "model_slim", "optim", "rng_state", "scaler", "scheduler", "step" ], "unsafe_globals": [ "numpy._core.multiarray._reconstruct", "numpy.dtype", "numpy.ndarray" ], "victim_weight_keys": 0 }, { "checkpoint_path": "runs/qwen-primary-full-t8-s1234/checkpoint/step_023000.pt", "checkpoint_sha256": "4cb3e0c7bfc7d010f152e60883e3ec1d0b6dc6b3a730b3f4b4d6c118f3a5dc79", "checkpoint_size_bytes": 98354880, "ema_payload_present": false, "hub_slug": "qwen-primary-full-t8-s1234", "model_state_prefixes": [ "feedback_proj", "init_token", "latent_pooler", "mode_embed", "out_ln", "token0_head", "transform" ], "optimizer_payload_present": false, "private_string_count": 0, "status": "PASS", "top_level_keys": [ "ckpt_mode", "ema_shadow", "model_slim", "optim", "rng_state", "scaler", "scheduler", "step" ], "unsafe_globals": [ "numpy._core.multiarray._reconstruct", "numpy.dtype", "numpy.ndarray" ], "victim_weight_keys": 0 }, { "checkpoint_path": "runs/qwen-primary-pooled-t8-s1234/checkpoint/step_024000.pt", "checkpoint_sha256": "555b319f453796b7471d47f4d0681fea25ce29a729ce5a19e3349b49b8737d29", "checkpoint_size_bytes": 98354880, "ema_payload_present": false, "hub_slug": "qwen-primary-pooled-t8-s1234", "model_state_prefixes": [ "feedback_proj", "init_token", "latent_pooler", "mode_embed", "out_ln", "token0_head", "transform" ], "optimizer_payload_present": false, "private_string_count": 0, "status": "PASS", "top_level_keys": [ "ckpt_mode", "ema_shadow", "model_slim", "optim", "rng_state", "scaler", "scheduler", "step" ], "unsafe_globals": [ "numpy._core.multiarray._reconstruct", "numpy.dtype", "numpy.ndarray" ], "victim_weight_keys": 0 } ], "status": "PASS", "torch_version": "2.12.0+cu130" }