| {"array_metadatas": [{"array_metadata": {"param_name": "step", "write_shape": [], "chunk_shape": [], "ext_metadata": null}}, {"array_metadata": {"param_name": "params.params.decoder.decoder_norm.scale", "write_shape": [192], "chunk_shape": [192], "ext_metadata": null}}, {"array_metadata": {"param_name": "params.params.decoder.layers.mlp.wi_0.kernel", "write_shape": [192, 32, 8192], "chunk_shape": [192, 32, 8192], "ext_metadata": null}}, {"array_metadata": {"param_name": "params.params.decoder.layers.mlp.wi_1.kernel", "write_shape": [192, 32, 8192], "chunk_shape": [192, 32, 8192], "ext_metadata": null}}, {"array_metadata": {"param_name": "params.params.decoder.layers.mlp.wo.kernel", "write_shape": [8192, 32, 192], "chunk_shape": [8192, 32, 192], "ext_metadata": null}}, {"array_metadata": {"param_name": "params.params.decoder.layers.post_self_attention_layer_norm.scale", "write_shape": [192, 32], "chunk_shape": [192, 32], "ext_metadata": null}}, {"array_metadata": {"param_name": "params.params.decoder.layers.pre_self_attention_layer_norm.scale", "write_shape": [192, 32], "chunk_shape": [192, 32], "ext_metadata": null}}, {"array_metadata": {"param_name": "params.params.decoder.layers.self_attention.key.kernel", "write_shape": [192, 32, 8, 96], "chunk_shape": [192, 32, 8, 96], "ext_metadata": null}}, {"array_metadata": {"param_name": "params.params.decoder.layers.self_attention.out.kernel", "write_shape": [32, 32, 96, 192], "chunk_shape": [32, 32, 96, 192], "ext_metadata": null}}, {"array_metadata": {"param_name": "params.params.decoder.layers.self_attention.query.kernel", "write_shape": [192, 32, 32, 96], "chunk_shape": [192, 32, 32, 96], "ext_metadata": null}}, {"array_metadata": {"param_name": "params.params.decoder.layers.self_attention.value.kernel", "write_shape": [192, 32, 8, 96], "chunk_shape": [192, 32, 8, 96], "ext_metadata": null}}, {"array_metadata": {"param_name": "params.params.decoder.logits_dense.kernel", "write_shape": [192, 50304], "chunk_shape": [192, 50304], "ext_metadata": null}}, {"array_metadata": {"param_name": "params.params.token_embedder.embedding", "write_shape": [50304, 192], "chunk_shape": [50304, 192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.count", "write_shape": [], "chunk_shape": [], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.mu.params.decoder.decoder_norm.scale", "write_shape": [192], "chunk_shape": [192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.mu.params.decoder.layers.mlp.wi_0.kernel", "write_shape": [192, 32, 8192], "chunk_shape": [192, 32, 8192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.mu.params.decoder.layers.mlp.wi_1.kernel", "write_shape": [192, 32, 8192], "chunk_shape": [192, 32, 8192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.mu.params.decoder.layers.mlp.wo.kernel", "write_shape": [8192, 32, 192], "chunk_shape": [8192, 32, 192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.mu.params.decoder.layers.post_self_attention_layer_norm.scale", "write_shape": [192, 32], "chunk_shape": [192, 32], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.mu.params.decoder.layers.pre_self_attention_layer_norm.scale", "write_shape": [192, 32], "chunk_shape": [192, 32], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.mu.params.decoder.layers.self_attention.key.kernel", "write_shape": [192, 32, 8, 96], "chunk_shape": [192, 32, 8, 96], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.mu.params.decoder.layers.self_attention.out.kernel", "write_shape": [32, 32, 96, 192], "chunk_shape": [32, 32, 96, 192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.mu.params.decoder.layers.self_attention.query.kernel", "write_shape": [192, 32, 32, 96], "chunk_shape": [192, 32, 32, 96], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.mu.params.decoder.layers.self_attention.value.kernel", "write_shape": [192, 32, 8, 96], "chunk_shape": [192, 32, 8, 96], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.mu.params.decoder.logits_dense.kernel", "write_shape": [192, 50304], "chunk_shape": [192, 50304], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.mu.params.token_embedder.embedding", "write_shape": [50304, 192], "chunk_shape": [50304, 192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.nu.params.decoder.decoder_norm.scale", "write_shape": [192], "chunk_shape": [192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.nu.params.decoder.layers.mlp.wi_0.kernel", "write_shape": [192, 32, 8192], "chunk_shape": [192, 32, 8192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.nu.params.decoder.layers.mlp.wi_1.kernel", "write_shape": [192, 32, 8192], "chunk_shape": [192, 32, 8192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.nu.params.decoder.layers.mlp.wo.kernel", "write_shape": [8192, 32, 192], "chunk_shape": [8192, 32, 192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.nu.params.decoder.layers.post_self_attention_layer_norm.scale", "write_shape": [192, 32], "chunk_shape": [192, 32], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.nu.params.decoder.layers.pre_self_attention_layer_norm.scale", "write_shape": [192, 32], "chunk_shape": [192, 32], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.nu.params.decoder.layers.self_attention.key.kernel", "write_shape": [192, 32, 8, 96], "chunk_shape": [192, 32, 8, 96], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.nu.params.decoder.layers.self_attention.out.kernel", "write_shape": [32, 32, 96, 192], "chunk_shape": [32, 32, 96, 192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.nu.params.decoder.layers.self_attention.query.kernel", "write_shape": [192, 32, 32, 96], "chunk_shape": [192, 32, 32, 96], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.nu.params.decoder.layers.self_attention.value.kernel", "write_shape": [192, 32, 8, 96], "chunk_shape": [192, 32, 8, 96], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.nu.params.decoder.logits_dense.kernel", "write_shape": [192, 50304], "chunk_shape": [192, 50304], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.0.nu.params.token_embedder.embedding", "write_shape": [50304, 192], "chunk_shape": [50304, 192], "ext_metadata": null}}, {"array_metadata": {"param_name": "opt_state.2.count", "write_shape": [], "chunk_shape": [], "ext_metadata": null}}]} |