Download cw-sim.json from adgodoyouk/atom-cw-sim: direct link, hf CLI and curl.
- Browser
- Download file 23.5 kB
-
https://huggingface.co/adgodoyouk/atom-cw-sim/resolve/main/cw-sim.json
- Command line
-
hf download hf://adgodoyouk/atom-cw-sim/cw-sim.json
-
curl -L -o cw-sim.json https://huggingface.co/adgodoyouk/atom-cw-sim/resolve/main/cw-sim.json
23.5 kB
| { | |
| "release_name": "cw-sim", | |
| "materialize_to": "results/cw_sim_base_run/final.pt", | |
| "config": { | |
| "filter_mode": "joint", | |
| "attempt_encoder": { | |
| "code_proj.weight": { | |
| "__tensor__": "/attempt_encoder/code_proj.weight" | |
| }, | |
| "code_proj.bias": { | |
| "__tensor__": "/attempt_encoder/code_proj.bias" | |
| }, | |
| "diff_proj.weight": { | |
| "__tensor__": "/attempt_encoder/diff_proj.weight" | |
| }, | |
| "diff_proj.bias": { | |
| "__tensor__": "/attempt_encoder/diff_proj.bias" | |
| }, | |
| "scalar_proj.weight": { | |
| "__tensor__": "/attempt_encoder/scalar_proj.weight" | |
| }, | |
| "scalar_proj.bias": { | |
| "__tensor__": "/attempt_encoder/scalar_proj.bias" | |
| }, | |
| "gru.weight_ih_l0": { | |
| "__tensor__": "/attempt_encoder/gru.weight_ih_l0" | |
| }, | |
| "gru.weight_hh_l0": { | |
| "__tensor__": "/attempt_encoder/gru.weight_hh_l0" | |
| }, | |
| "gru.bias_ih_l0": { | |
| "__tensor__": "/attempt_encoder/gru.bias_ih_l0" | |
| }, | |
| "gru.bias_hh_l0": { | |
| "__tensor__": "/attempt_encoder/gru.bias_hh_l0" | |
| }, | |
| "summary_head.weight": { | |
| "__tensor__": "/attempt_encoder/summary_head.weight" | |
| }, | |
| "summary_head.bias": { | |
| "__tensor__": "/attempt_encoder/summary_head.bias" | |
| } | |
| }, | |
| "filt": { | |
| "episode_encoder.item_proj.net.0.weight": { | |
| "__tensor__": "/filt/episode_encoder.item_proj.net.0.weight" | |
| }, | |
| "episode_encoder.item_proj.net.0.bias": { | |
| "__tensor__": "/filt/episode_encoder.item_proj.net.0.bias" | |
| }, | |
| "episode_encoder.item_proj.net.2.weight": { | |
| "__tensor__": "/filt/episode_encoder.item_proj.net.2.weight" | |
| }, | |
| "episode_encoder.item_proj.net.2.bias": { | |
| "__tensor__": "/filt/episode_encoder.item_proj.net.2.bias" | |
| }, | |
| "episode_encoder.prev_proj.weight": { | |
| "__tensor__": "/filt/episode_encoder.prev_proj.weight" | |
| }, | |
| "episode_encoder.prev_proj.bias": { | |
| "__tensor__": "/filt/episode_encoder.prev_proj.bias" | |
| }, | |
| "knowledge_filter.gru.weight_ih_l0": { | |
| "__tensor__": "/filt/knowledge_filter.gru.weight_ih_l0" | |
| }, | |
| "knowledge_filter.gru.weight_hh_l0": { | |
| "__tensor__": "/filt/knowledge_filter.gru.weight_hh_l0" | |
| }, | |
| "knowledge_filter.gru.bias_ih_l0": { | |
| "__tensor__": "/filt/knowledge_filter.gru.bias_ih_l0" | |
| }, | |
| "knowledge_filter.gru.bias_hh_l0": { | |
| "__tensor__": "/filt/knowledge_filter.gru.bias_hh_l0" | |
| }, | |
| "knowledge_filter.mu_head.weight": { | |
| "__tensor__": "/filt/knowledge_filter.mu_head.weight" | |
| }, | |
| "knowledge_filter.mu_head.bias": { | |
| "__tensor__": "/filt/knowledge_filter.mu_head.bias" | |
| }, | |
| "knowledge_filter.precision_head.weight": { | |
| "__tensor__": "/filt/knowledge_filter.precision_head.weight" | |
| }, | |
| "knowledge_filter.precision_head.bias": { | |
| "__tensor__": "/filt/knowledge_filter.precision_head.bias" | |
| } | |
| }, | |
| "outcome_head": { | |
| "trunk.net.0.weight": { | |
| "__tensor__": "/outcome_head/trunk.net.0.weight" | |
| }, | |
| "trunk.net.0.bias": { | |
| "__tensor__": "/outcome_head/trunk.net.0.bias" | |
| }, | |
| "trunk.net.2.weight": { | |
| "__tensor__": "/outcome_head/trunk.net.2.weight" | |
| }, | |
| "trunk.net.2.bias": { | |
| "__tensor__": "/outcome_head/trunk.net.2.bias" | |
| }, | |
| "prefix_proj.weight": { | |
| "__tensor__": "/outcome_head/prefix_proj.weight" | |
| }, | |
| "prefix_proj.bias": { | |
| "__tensor__": "/outcome_head/prefix_proj.bias" | |
| }, | |
| "gamma.weight": { | |
| "__tensor__": "/outcome_head/gamma.weight" | |
| }, | |
| "gamma.bias": { | |
| "__tensor__": "/outcome_head/gamma.bias" | |
| }, | |
| "beta.weight": { | |
| "__tensor__": "/outcome_head/beta.weight" | |
| }, | |
| "beta.bias": { | |
| "__tensor__": "/outcome_head/beta.bias" | |
| }, | |
| "outcome.weight": { | |
| "__tensor__": "/outcome_head/outcome.weight" | |
| }, | |
| "outcome.bias": { | |
| "__tensor__": "/outcome_head/outcome.bias" | |
| } | |
| }, | |
| "test_vector_head": { | |
| "trunk.net.0.weight": { | |
| "__tensor__": "/test_vector_head/trunk.net.0.weight" | |
| }, | |
| "trunk.net.0.bias": { | |
| "__tensor__": "/test_vector_head/trunk.net.0.bias" | |
| }, | |
| "trunk.net.2.weight": { | |
| "__tensor__": "/test_vector_head/trunk.net.2.weight" | |
| }, | |
| "trunk.net.2.bias": { | |
| "__tensor__": "/test_vector_head/trunk.net.2.bias" | |
| }, | |
| "prefix_proj.weight": { | |
| "__tensor__": "/test_vector_head/prefix_proj.weight" | |
| }, | |
| "prefix_proj.bias": { | |
| "__tensor__": "/test_vector_head/prefix_proj.bias" | |
| }, | |
| "gamma.weight": { | |
| "__tensor__": "/test_vector_head/gamma.weight" | |
| }, | |
| "gamma.bias": { | |
| "__tensor__": "/test_vector_head/gamma.bias" | |
| }, | |
| "beta.weight": { | |
| "__tensor__": "/test_vector_head/beta.weight" | |
| }, | |
| "beta.bias": { | |
| "__tensor__": "/test_vector_head/beta.bias" | |
| }, | |
| "outcome.weight": { | |
| "__tensor__": "/test_vector_head/outcome.weight" | |
| }, | |
| "outcome.bias": { | |
| "__tensor__": "/test_vector_head/outcome.bias" | |
| } | |
| }, | |
| "hyper": { | |
| "net.0.weight": { | |
| "__tensor__": "/hyper/net.0.weight" | |
| }, | |
| "net.0.bias": { | |
| "__tensor__": "/hyper/net.0.bias" | |
| }, | |
| "net.2.weight": { | |
| "__tensor__": "/hyper/net.2.weight" | |
| }, | |
| "net.2.bias": { | |
| "__tensor__": "/hyper/net.2.bias" | |
| } | |
| }, | |
| "basis": { | |
| "model.language_model.layers.0.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.0.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.0.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.0.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.0.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.0.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.1.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.1.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.1.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.1.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.1.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.1.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.2.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.2.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.2.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.2.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.2.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.2.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.3.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.3.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.3.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.3.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.3.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.3.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.4.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.4.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.4.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.4.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.4.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.4.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.5.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.5.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.5.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.5.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.5.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.5.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.6.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.6.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.6.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.6.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.6.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.6.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.7.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.7.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.7.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.7.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.7.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.7.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.8.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.8.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.8.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.8.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.8.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.8.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.9.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.9.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.9.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.9.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.9.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.9.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.10.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.10.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.10.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.10.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.10.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.10.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.11.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.11.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.11.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.11.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.11.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.11.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.12.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.12.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.12.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.12.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.12.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.12.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.13.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.13.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.13.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.13.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.13.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.13.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.14.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.14.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.14.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.14.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.14.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.14.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.15.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.15.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.15.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.15.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.15.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.15.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.16.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.16.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.16.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.16.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.16.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.16.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.17.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.17.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.17.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.17.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.17.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.17.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.18.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.18.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.18.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.18.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.18.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.18.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.19.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.19.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.19.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.19.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.19.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.19.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.20.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.20.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.20.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.20.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.20.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.20.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.21.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.21.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.21.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.21.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.21.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.21.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.22.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.22.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.22.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.22.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.22.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.22.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.23.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.23.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.23.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.23.self_attn.v_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.23.self_attn.v_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.23.self_attn.v_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.24.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.24.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.24.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.25.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.25.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.25.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.26.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.26.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.26.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.27.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.27.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.27.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.28.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.28.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.28.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.29.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.29.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.29.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.30.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.30.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.30.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.31.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.31.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.31.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.32.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.32.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.32.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.33.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.33.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.33.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.34.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.34.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.34.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.35.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.35.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.35.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.36.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.36.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.36.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.37.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.37.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.37.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.38.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.38.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.38.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.39.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.39.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.39.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.40.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.40.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.40.self_attn.q_proj/B" | |
| } | |
| }, | |
| "model.language_model.layers.41.self_attn.q_proj": { | |
| "A": { | |
| "__tensor__": "/basis/model.language_model.layers.41.self_attn.q_proj/A" | |
| }, | |
| "B": { | |
| "__tensor__": "/basis/model.language_model.layers.41.self_attn.q_proj/B" | |
| } | |
| } | |
| }, | |
| "architecture": { | |
| "latent_dim": 32, | |
| "episode_hidden": 64, | |
| "filter_hidden": 64, | |
| "attempt_proj_dim": 32, | |
| "attempt_hidden": 64, | |
| "summary_dim": 16, | |
| "codec_dim": 64 | |
| }, | |
| "token_loss_weight": 1.0, | |
| "outcome_loss_weight": 0.3, | |
| "test_vector_loss_weight": 0.1, | |
| "ratchet_checkpoint_sha256": "8d22b6914e041dfa88bf911b1a498fa0c87af30488b07f6e49a72adca731c544", | |
| "learner_split_sha256": "c6b8eb8c45e3ff9313a02493a40966a5ec98a09f2780a5adebfb0fc6b26b7237", | |
| "git_sha": "unknown", | |
| "seed": 0, | |
| "sampler_semantics": "reparameterize_from_log_sigma_v1_2026-08-15", | |
| "step": 341, | |
| "max_test_cases": 32, | |
| "held_out_items": [ | |
| 51, | |
| 56, | |
| 57, | |
| 64, | |
| 70, | |
| 112, | |
| 118 | |
| ], | |
| "codec_path": "transforms/problem_mean_L15.npz", | |
| "train_regime": "cw_sim_full_run.py --filter-mode joint", | |
| "epochs": 1.9941520467836258, | |
| "headline_metrics": { | |
| "arms": { | |
| "personalised": { | |
| "n": 18158, | |
| "token_nll_mean": 0.148970365524292, | |
| "exact_match_rate": 0.15579909086227417, | |
| "top1_rate": 0.9604315161705017 | |
| }, | |
| "dw0": { | |
| "n": 18158, | |
| "token_nll_mean": 0.34834980964660645, | |
| "exact_match_rate": 0.0, | |
| "top1_rate": 0.9222095012664795 | |
| }, | |
| "shuffled_u": { | |
| "n": 18000, | |
| "token_nll_mean": 0.14966580271720886, | |
| "exact_match_rate": 0.15549999475479126, | |
| "top1_rate": 0.9602310061454773 | |
| }, | |
| "u_mean": { | |
| "n": 18158, | |
| "token_nll_mean": 0.15335048735141754, | |
| "exact_match_rate": 0.14902521669864655, | |
| "top1_rate": 0.9597863554954529 | |
| } | |
| }, | |
| "personalisation_gap_token_nll": -0.0006954371929168701, | |
| "personalisation_gap_ci": { | |
| "point_delta": -0.000721275806427002, | |
| "ci95": [ | |
| -0.0009116943925619126, | |
| -0.0005604498088359834 | |
| ], | |
| "excludes_zero": true, | |
| "n_boot": 2000, | |
| "n_learners": 110, | |
| "n_rows": 18000, | |
| "n_rows_intersection": 18000 | |
| } | |
| }, | |
| "canary9_belief_dependence": null | |
| } | |
| } |