diff --git a/.gitattributes b/.gitattributes index 5ddd97a0e76fa52fb9cc8b51a7c8492c5b7921fe..ed625dd5a761504881d178fc115c1cc40e7ce2b7 100644 --- a/.gitattributes +++ b/.gitattributes @@ -35,3 +35,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text *tfevents* filter=lfs diff=lfs merge=lfs -text models/gemma-3-270m-it-wq4k-ef16/tokenizer.json filter=lfs diff=lfs merge=lfs -text models/google-embeddinggemma-300m-wq4k-ef16/tokenizer.json filter=lfs diff=lfs merge=lfs -text +models/translategemma-4b-it-wq4k-ef16-hf16/tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/manifest.json b/models/translategemma-4b-it-wq4k-ef16-hf16/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..998e56691cd5125f49dd06e5c6295e82e265ed30 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/manifest.json @@ -0,0 +1,6464 @@ +{ + "version": 1, + "modelId": "translategemma-4b-it-wq4k-ef16-hf16", + "modelType": "transformer", + "quantization": "Q4_K_M", + "quantizationInfo": { + "weights": "q4k", + "embeddings": "f16", + "compute": "f16", + "layout": "row", + "variantTag": "wq4k-ef16" + }, + "architecture": { + "numLayers": 34, + "hiddenSize": 2560, + "intermediateSize": 10240, + "numAttentionHeads": 8, + "numKeyValueHeads": 4, + "headDim": 256, + "vocabSize": 262208, + "maxSeqLen": 131072, + "ropeTheta": 1000000 + }, + "moeConfig": null, + "inference": { + "schema": "doppler.execution/v0", + "presetId": "translategemma", + "attention": { + "queryPreAttnScalar": 16, + "attnLogitSoftcapping": null, + "slidingWindow": null, + "queryKeyNorm": true, + "attentionOutputGate": false, + "causal": true, + "attentionBias": false + }, + "normalization": { + "rmsNormEps": 0.000001, + "rmsNormWeightOffset": true, + "postAttentionNorm": true, + "preFeedforwardNorm": true, + "postFeedforwardNorm": true + }, + "ffn": { + "activation": "gelu", + "gatedActivation": true, + "swigluLimit": null + }, + "rope": { + "ropeTheta": 1000000, + "ropeLocalTheta": 10000, + "ropeScalingType": null, + "ropeScalingFactor": 1, + "yarnBetaFast": null, + "yarnBetaSlow": null, + "yarnOriginalMaxPos": null, + "ropeLocalScalingType": null, + "ropeLocalScalingFactor": 1, + "ropeLocalYarnBetaFast": null, + "ropeLocalYarnBetaSlow": null, + "ropeLocalYarnOriginalMaxPos": null + }, + "output": { + "finalLogitSoftcapping": null, + "tieWordEmbeddings": true, + "scaleEmbeddings": true, + "embeddingTranspose": false, + "embeddingVocabSize": null + }, + "layerPattern": { + "type": "every_n", + "globalPattern": null, + "period": 6, + "offset": 5, + "layerTypes": null + }, + "chatTemplate": { + "type": "translategemma", + "enabled": true + }, + "pipeline": null, + "sessionDefaults": { + "compute": { + "defaults": { + "activationDtype": "f16", + "mathDtype": "f16", + "accumDtype": "f32", + "outputDtype": "f16" + }, + "kernelProfiles": [ + { + "kernelRef": { + "id": "gather.f16.main", + "version": "1.0.0", + "digest": "sha256:a4829f4067091c98ad6ebbc9b0744cdd5bbcd4fbf6092b2f7cc7f1098695860f" + } + }, + { + "kernelRef": { + "id": "rmsnorm.f16.main", + "version": "1.0.0", + "digest": "sha256:21d6d702cfcb8d653d8e105bbbc522e85704fa5628b865a7d36cde338ca5c779" + } + }, + { + "kernelRef": { + "id": "matmul.gemv.subgroup.f16a.main.vec4", + "version": "1.0.0", + "digest": "sha256:f227a403cdf9717dd68224c9ea55708ffe14c618d8146f5d48b42af0f253df29" + } + }, + { + "kernelRef": { + "id": "rope.f16.main", + "version": "1.0.0", + "digest": "sha256:a87f2964b77e851a2fbcc88305adeecaf8eba372291d83a71b817c8ef3da5c58" + } + }, + { + "kernelRef": { + "id": "attention.decode.online.f16.main", + "version": "1.0.0", + "digest": "sha256:953acecd07bd1e50829f26a7e6abb6c653331504d23914b77538da99ee1474dc" + } + }, + { + "kernelRef": { + "id": "residual.f16.main", + "version": "1.0.0", + "digest": "sha256:d392433f3065d1caf68b033219f4ffacf022dc1f90fc3cf3fd620e4ba49f3219" + } + }, + { + "kernelRef": { + "id": "gelu.f16.main", + "version": "1.0.0", + "digest": "sha256:34b2f981e8b96bd624748b84bad0512940a01ada1721e0ffb8e149259b82b22a" + } + }, + { + "kernelRef": { + "id": "matmul.f16.main", + "version": "1.0.0", + "digest": "sha256:b2a7e4a4db27d008da957225b361ef688f299ece792c844e2d2542b21d3fdfbe" + } + }, + { + "kernelRef": { + "id": "attention.streaming.f16.main", + "version": "1.0.0", + "digest": "sha256:2a1591412e581fb855bc54003c616819008c9ee21b56adee81afbe672eba32c9" + } + }, + { + "kernelRef": { + "id": "matmul.gemv.subgroup.f16a.main.multicol", + "version": "1.0.0", + "digest": "sha256:c8e86ecbbefa27a3b7366af676d89a992c2e951329cdf19abb57b9c90144379e" + } + }, + { + "kernelRef": { + "id": "sample.f16.sample.single.pass", + "version": "1.0.0", + "digest": "sha256:f3551288786f0a47f5222e07ab7f54a646e422c2493a62b8074202601eb06ddb" + } + } + ] + }, + "kvcache": { + "layout": "paged", + "kvDtype": "f16", + "pageSize": 128, + "windowSize": 1024, + "bdpaVocabSize": 4096 + }, + "decodeLoop": { + "batchSize": 16, + "stopCheckMode": "batch", + "readbackInterval": 8, + "ringTokens": 8, + "ringStop": 8, + "ringStaging": 8 + } + }, + "execution": { + "steps": [ + { + "id": "embed_tokens", + "phase": "both", + "section": "preLayer", + "op": "embed", + "kernel": "gather_f16.wgsl", + "entry": "main", + "weights": "embed_tokens", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "gather.f16.main", + "version": "1.0.0", + "digest": "sha256:a4829f4067091c98ad6ebbc9b0744cdd5bbcd4fbf6092b2f7cc7f1098695860f" + } + }, + { + "id": "decode_input_norm", + "phase": "decode", + "section": "layer", + "op": "input_norm", + "kernel": "rmsnorm_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "rmsnorm.f16.main", + "version": "1.0.0", + "digest": "sha256:21d6d702cfcb8d653d8e105bbbc522e85704fa5628b865a7d36cde338ca5c779" + } + }, + { + "id": "decode_q_proj", + "phase": "decode", + "section": "layer", + "op": "q_proj", + "kernel": "matmul_gemv_subgroup_f16a.wgsl", + "entry": "main_vec4", + "weights": "layer.{L}.self_attn.q_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.gemv.subgroup.f16a.main.vec4", + "version": "1.0.0", + "digest": "sha256:f227a403cdf9717dd68224c9ea55708ffe14c618d8146f5d48b42af0f253df29" + } + }, + { + "id": "decode_k_proj", + "phase": "decode", + "section": "layer", + "op": "k_proj", + "kernel": "matmul_gemv_subgroup_f16a.wgsl", + "entry": "main_vec4", + "weights": "layer.{L}.self_attn.k_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.gemv.subgroup.f16a.main.vec4", + "version": "1.0.0", + "digest": "sha256:f227a403cdf9717dd68224c9ea55708ffe14c618d8146f5d48b42af0f253df29" + } + }, + { + "id": "decode_v_proj", + "phase": "decode", + "section": "layer", + "op": "v_proj", + "kernel": "matmul_gemv_subgroup_f16a.wgsl", + "entry": "main_vec4", + "weights": "layer.{L}.self_attn.v_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.gemv.subgroup.f16a.main.vec4", + "version": "1.0.0", + "digest": "sha256:f227a403cdf9717dd68224c9ea55708ffe14c618d8146f5d48b42af0f253df29" + } + }, + { + "id": "decode_rope_q", + "phase": "decode", + "section": "layer", + "op": "rope_q", + "kernel": "rope_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "rope.f16.main", + "version": "1.0.0", + "digest": "sha256:a87f2964b77e851a2fbcc88305adeecaf8eba372291d83a71b817c8ef3da5c58" + } + }, + { + "id": "decode_rope_k", + "phase": "decode", + "section": "layer", + "op": "rope_k", + "kernel": "rope_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "rope.f16.main", + "version": "1.0.0", + "digest": "sha256:a87f2964b77e851a2fbcc88305adeecaf8eba372291d83a71b817c8ef3da5c58" + } + }, + { + "id": "decode_attention", + "phase": "decode", + "section": "layer", + "op": "attention", + "kernel": "attention_decode_online_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "attention.decode.online.f16.main", + "version": "1.0.0", + "digest": "sha256:953acecd07bd1e50829f26a7e6abb6c653331504d23914b77538da99ee1474dc" + } + }, + { + "id": "decode_o_proj", + "phase": "decode", + "section": "layer", + "op": "o_proj", + "kernel": "matmul_gemv_subgroup_f16a.wgsl", + "entry": "main_vec4", + "weights": "layer.{L}.self_attn.o_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.gemv.subgroup.f16a.main.vec4", + "version": "1.0.0", + "digest": "sha256:f227a403cdf9717dd68224c9ea55708ffe14c618d8146f5d48b42af0f253df29" + } + }, + { + "id": "decode_attn_residual", + "phase": "decode", + "section": "layer", + "op": "attn_residual", + "kernel": "residual_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "residual.f16.main", + "version": "1.0.0", + "digest": "sha256:d392433f3065d1caf68b033219f4ffacf022dc1f90fc3cf3fd620e4ba49f3219" + } + }, + { + "id": "decode_post_attn_norm", + "phase": "decode", + "section": "layer", + "op": "post_attn_norm", + "kernel": "rmsnorm_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "rmsnorm.f16.main", + "version": "1.0.0", + "digest": "sha256:21d6d702cfcb8d653d8e105bbbc522e85704fa5628b865a7d36cde338ca5c779" + } + }, + { + "id": "decode_gate_proj", + "phase": "decode", + "section": "layer", + "op": "gate_proj", + "kernel": "matmul_gemv_subgroup_f16a.wgsl", + "entry": "main_vec4", + "weights": "layer.{L}.mlp.gate_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.gemv.subgroup.f16a.main.vec4", + "version": "1.0.0", + "digest": "sha256:f227a403cdf9717dd68224c9ea55708ffe14c618d8146f5d48b42af0f253df29" + } + }, + { + "id": "decode_up_proj", + "phase": "decode", + "section": "layer", + "op": "up_proj", + "kernel": "matmul_gemv_subgroup_f16a.wgsl", + "entry": "main_vec4", + "weights": "layer.{L}.mlp.up_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.gemv.subgroup.f16a.main.vec4", + "version": "1.0.0", + "digest": "sha256:f227a403cdf9717dd68224c9ea55708ffe14c618d8146f5d48b42af0f253df29" + } + }, + { + "id": "decode_activation", + "phase": "decode", + "section": "layer", + "op": "activation", + "kernel": "gelu_f16.wgsl", + "entry": "main", + "constants": { + "HAS_GATE": true + }, + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "gelu.f16.main", + "version": "1.0.0", + "digest": "sha256:34b2f981e8b96bd624748b84bad0512940a01ada1721e0ffb8e149259b82b22a" + } + }, + { + "id": "decode_down_proj", + "phase": "decode", + "section": "layer", + "op": "down_proj", + "kernel": "matmul_gemv_subgroup_f16a.wgsl", + "entry": "main_vec4", + "weights": "layer.{L}.mlp.down_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.gemv.subgroup.f16a.main.vec4", + "version": "1.0.0", + "digest": "sha256:f227a403cdf9717dd68224c9ea55708ffe14c618d8146f5d48b42af0f253df29" + } + }, + { + "id": "decode_ffn_residual", + "phase": "decode", + "section": "layer", + "op": "ffn_residual", + "kernel": "residual_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "residual.f16.main", + "version": "1.0.0", + "digest": "sha256:d392433f3065d1caf68b033219f4ffacf022dc1f90fc3cf3fd620e4ba49f3219" + } + }, + { + "id": "prefill_input_norm", + "phase": "prefill", + "section": "layer", + "op": "input_norm", + "kernel": "rmsnorm_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "rmsnorm.f16.main", + "version": "1.0.0", + "digest": "sha256:21d6d702cfcb8d653d8e105bbbc522e85704fa5628b865a7d36cde338ca5c779" + } + }, + { + "id": "prefill_q_proj", + "phase": "prefill", + "section": "layer", + "op": "q_proj", + "kernel": "matmul_f16.wgsl", + "entry": "main", + "weights": "layer.{L}.self_attn.q_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.f16.main", + "version": "1.0.0", + "digest": "sha256:b2a7e4a4db27d008da957225b361ef688f299ece792c844e2d2542b21d3fdfbe" + } + }, + { + "id": "prefill_k_proj", + "phase": "prefill", + "section": "layer", + "op": "k_proj", + "kernel": "matmul_f16.wgsl", + "entry": "main", + "weights": "layer.{L}.self_attn.k_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.f16.main", + "version": "1.0.0", + "digest": "sha256:b2a7e4a4db27d008da957225b361ef688f299ece792c844e2d2542b21d3fdfbe" + } + }, + { + "id": "prefill_v_proj", + "phase": "prefill", + "section": "layer", + "op": "v_proj", + "kernel": "matmul_f16.wgsl", + "entry": "main", + "weights": "layer.{L}.self_attn.v_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.f16.main", + "version": "1.0.0", + "digest": "sha256:b2a7e4a4db27d008da957225b361ef688f299ece792c844e2d2542b21d3fdfbe" + } + }, + { + "id": "prefill_rope_q", + "phase": "prefill", + "section": "layer", + "op": "rope_q", + "kernel": "rope_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "rope.f16.main", + "version": "1.0.0", + "digest": "sha256:a87f2964b77e851a2fbcc88305adeecaf8eba372291d83a71b817c8ef3da5c58" + } + }, + { + "id": "prefill_rope_k", + "phase": "prefill", + "section": "layer", + "op": "rope_k", + "kernel": "rope_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "rope.f16.main", + "version": "1.0.0", + "digest": "sha256:a87f2964b77e851a2fbcc88305adeecaf8eba372291d83a71b817c8ef3da5c58" + } + }, + { + "id": "prefill_attention", + "phase": "prefill", + "section": "layer", + "op": "attention", + "kernel": "attention_streaming_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "attention.streaming.f16.main", + "version": "1.0.0", + "digest": "sha256:2a1591412e581fb855bc54003c616819008c9ee21b56adee81afbe672eba32c9" + } + }, + { + "id": "prefill_o_proj", + "phase": "prefill", + "section": "layer", + "op": "o_proj", + "kernel": "matmul_f16.wgsl", + "entry": "main", + "weights": "layer.{L}.self_attn.o_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.f16.main", + "version": "1.0.0", + "digest": "sha256:b2a7e4a4db27d008da957225b361ef688f299ece792c844e2d2542b21d3fdfbe" + } + }, + { + "id": "prefill_attn_residual", + "phase": "prefill", + "section": "layer", + "op": "attn_residual", + "kernel": "residual_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "residual.f16.main", + "version": "1.0.0", + "digest": "sha256:d392433f3065d1caf68b033219f4ffacf022dc1f90fc3cf3fd620e4ba49f3219" + } + }, + { + "id": "prefill_post_attn_norm", + "phase": "prefill", + "section": "layer", + "op": "post_attn_norm", + "kernel": "rmsnorm_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "rmsnorm.f16.main", + "version": "1.0.0", + "digest": "sha256:21d6d702cfcb8d653d8e105bbbc522e85704fa5628b865a7d36cde338ca5c779" + } + }, + { + "id": "prefill_gate_proj", + "phase": "prefill", + "section": "layer", + "op": "gate_proj", + "kernel": "matmul_f16.wgsl", + "entry": "main", + "weights": "layer.{L}.mlp.gate_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.f16.main", + "version": "1.0.0", + "digest": "sha256:b2a7e4a4db27d008da957225b361ef688f299ece792c844e2d2542b21d3fdfbe" + } + }, + { + "id": "prefill_up_proj", + "phase": "prefill", + "section": "layer", + "op": "up_proj", + "kernel": "matmul_f16.wgsl", + "entry": "main", + "weights": "layer.{L}.mlp.up_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.f16.main", + "version": "1.0.0", + "digest": "sha256:b2a7e4a4db27d008da957225b361ef688f299ece792c844e2d2542b21d3fdfbe" + } + }, + { + "id": "prefill_activation", + "phase": "prefill", + "section": "layer", + "op": "activation", + "kernel": "gelu_f16.wgsl", + "entry": "main", + "constants": { + "HAS_GATE": true + }, + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "gelu.f16.main", + "version": "1.0.0", + "digest": "sha256:34b2f981e8b96bd624748b84bad0512940a01ada1721e0ffb8e149259b82b22a" + } + }, + { + "id": "prefill_down_proj", + "phase": "prefill", + "section": "layer", + "op": "down_proj", + "kernel": "matmul_f16.wgsl", + "entry": "main", + "weights": "layer.{L}.mlp.down_proj", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.f16.main", + "version": "1.0.0", + "digest": "sha256:b2a7e4a4db27d008da957225b361ef688f299ece792c844e2d2542b21d3fdfbe" + } + }, + { + "id": "prefill_ffn_residual", + "phase": "prefill", + "section": "layer", + "op": "ffn_residual", + "kernel": "residual_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "residual.f16.main", + "version": "1.0.0", + "digest": "sha256:d392433f3065d1caf68b033219f4ffacf022dc1f90fc3cf3fd620e4ba49f3219" + } + }, + { + "id": "final_norm", + "phase": "both", + "section": "postLayer", + "op": "final_norm", + "kernel": "rmsnorm_f16.wgsl", + "entry": "main", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "rmsnorm.f16.main", + "version": "1.0.0", + "digest": "sha256:21d6d702cfcb8d653d8e105bbbc522e85704fa5628b865a7d36cde338ca5c779" + } + }, + { + "id": "decode_lm_head", + "phase": "decode", + "section": "postLayer", + "op": "lm_head", + "kernel": "matmul_gemv_subgroup_f16a.wgsl", + "entry": "main_multicol", + "weights": "lm_head", + "constants": { + "MULTICOL_COLS_PER_WG": 64, + "MULTICOL_THREADS_PER_COL": 4 + }, + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.gemv.subgroup.f16a.main.multicol", + "version": "1.0.0", + "digest": "sha256:c8e86ecbbefa27a3b7366af676d89a992c2e951329cdf19abb57b9c90144379e" + } + }, + { + "id": "prefill_lm_head", + "phase": "prefill", + "section": "postLayer", + "op": "lm_head_prefill", + "kernel": "matmul_f16.wgsl", + "entry": "main", + "weights": "lm_head", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "matmul.f16.main", + "version": "1.0.0", + "digest": "sha256:b2a7e4a4db27d008da957225b361ef688f299ece792c844e2d2542b21d3fdfbe" + } + }, + { + "id": "decode_sample", + "phase": "decode", + "section": "sampling", + "op": "sample", + "kernel": "sample_f16.wgsl", + "entry": "sample_single_pass", + "layers": "all", + "src": "state", + "dst": "state", + "kernelRef": { + "id": "sample.f16.sample.single.pass", + "version": "1.0.0", + "digest": "sha256:f3551288786f0a47f5222e07ab7f54a646e422c2493a62b8074202601eb06ddb" + } + } + ], + "policies": { + "precisionPrecedence": "step_then_kernel_profile_then_session_default", + "unsupportedPrecision": "error", + "dtypeTransition": "require_cast_step", + "unresolvedKernel": "error" + } + }, + "defaultKernelPath": "gemma3-q4k-dequant-f16a-online" + }, + "shards": [ + { + "index": 0, + "filename": "shard_00000.bin", + "size": 67108864, + "hash": "411f49f08d7bb17643003ee332adc2d1c4fd6c6e98b3e5cde8608344a1620bef", + "blake3": "411f49f08d7bb17643003ee332adc2d1c4fd6c6e98b3e5cde8608344a1620bef", + "offset": 0 + }, + { + "index": 1, + "filename": "shard_00001.bin", + "size": 67108864, + "hash": "fba3872816acd84b58cda34eafa157f1b779abc64752e6ea13c81854285a4340", + "blake3": "fba3872816acd84b58cda34eafa157f1b779abc64752e6ea13c81854285a4340", + "offset": 67108864 + }, + { + "index": 2, + "filename": "shard_00002.bin", + "size": 67108864, + "hash": "54ebf5c4e5d012c2d36f77935d128a509b77848745c66d3e3ecef80111ea5ca3", + "blake3": "54ebf5c4e5d012c2d36f77935d128a509b77848745c66d3e3ecef80111ea5ca3", + "offset": 134217728 + }, + { + "index": 3, + "filename": "shard_00003.bin", + "size": 67108864, + "hash": "ccdde5b606be302fe5e48741d2eddc351daa62848b22a83edf8bd2143d83b59b", + "blake3": "ccdde5b606be302fe5e48741d2eddc351daa62848b22a83edf8bd2143d83b59b", + "offset": 201326592 + }, + { + "index": 4, + "filename": "shard_00004.bin", + "size": 67108864, + "hash": "400a0c2d8758c589eaca83dbea91c010f627775b56a49a278474c23b3abe074e", + "blake3": "400a0c2d8758c589eaca83dbea91c010f627775b56a49a278474c23b3abe074e", + "offset": 268435456 + }, + { + "index": 5, + "filename": "shard_00005.bin", + "size": 67108864, + "hash": "4ec6053cc07478d4914efc7c08946ef36bfbf76664fa09e50429adf2fb81e870", + "blake3": "4ec6053cc07478d4914efc7c08946ef36bfbf76664fa09e50429adf2fb81e870", + "offset": 335544320 + }, + { + "index": 6, + "filename": "shard_00006.bin", + "size": 67108864, + "hash": "ffbb43ee273ca222a4b6d44010e8a97bdbe5e08a9fe022afd7e54c5a6235845a", + "blake3": "ffbb43ee273ca222a4b6d44010e8a97bdbe5e08a9fe022afd7e54c5a6235845a", + "offset": 402653184 + }, + { + "index": 7, + "filename": "shard_00007.bin", + "size": 67108864, + "hash": "6c15fb2215d67e35a8ff26c8a545dc99f1ba387c4dfc829aafe7ee8772f89c6c", + "blake3": "6c15fb2215d67e35a8ff26c8a545dc99f1ba387c4dfc829aafe7ee8772f89c6c", + "offset": 469762048 + }, + { + "index": 8, + "filename": "shard_00008.bin", + "size": 67108864, + "hash": "0f841f6fb9c687b56b65557e8b2a1e8f963dd3e6dd51e2d55f306a4d21a3fc5d", + "blake3": "0f841f6fb9c687b56b65557e8b2a1e8f963dd3e6dd51e2d55f306a4d21a3fc5d", + "offset": 536870912 + }, + { + "index": 9, + "filename": "shard_00009.bin", + "size": 67108864, + "hash": "022451c31d67ac4ed12df19d4a29b18f8eb7c1f1173030ea45f381b2b65c4f1f", + "blake3": "022451c31d67ac4ed12df19d4a29b18f8eb7c1f1173030ea45f381b2b65c4f1f", + "offset": 603979776 + }, + { + "index": 10, + "filename": "shard_00010.bin", + "size": 67108864, + "hash": "8be8327a6bf48d480a205a0592fd99bb46d077e473e98a40ece6d592d5e27508", + "blake3": "8be8327a6bf48d480a205a0592fd99bb46d077e473e98a40ece6d592d5e27508", + "offset": 671088640 + }, + { + "index": 11, + "filename": "shard_00011.bin", + "size": 67108864, + "hash": "4d59a9615c7d79107bcd826564c5ceb6d06e9ae0f63a5150496818183933692f", + "blake3": "4d59a9615c7d79107bcd826564c5ceb6d06e9ae0f63a5150496818183933692f", + "offset": 738197504 + }, + { + "index": 12, + "filename": "shard_00012.bin", + "size": 67108864, + "hash": "73021f8a130aa7f04e95c28cd8dd21a8a352462b8e0738a5cf0fc5526352c82f", + "blake3": "73021f8a130aa7f04e95c28cd8dd21a8a352462b8e0738a5cf0fc5526352c82f", + "offset": 805306368 + }, + { + "index": 13, + "filename": "shard_00013.bin", + "size": 67108864, + "hash": "cbfe08f3c87de59483784b6d72fbba67bf7b72b870ad70400c649c018ce98aaf", + "blake3": "cbfe08f3c87de59483784b6d72fbba67bf7b72b870ad70400c649c018ce98aaf", + "offset": 872415232 + }, + { + "index": 14, + "filename": "shard_00014.bin", + "size": 67108864, + "hash": "d763c90ee261dd106d37d2c3214fe5c4b80c2d6990c582cc9ad7fe186bda896d", + "blake3": "d763c90ee261dd106d37d2c3214fe5c4b80c2d6990c582cc9ad7fe186bda896d", + "offset": 939524096 + }, + { + "index": 15, + "filename": "shard_00015.bin", + "size": 67108864, + "hash": "e985a59e76af71e32a88a902a8ccc12ecf52bbfb5a1ec84112bd4c1a2e902178", + "blake3": "e985a59e76af71e32a88a902a8ccc12ecf52bbfb5a1ec84112bd4c1a2e902178", + "offset": 1006632960 + }, + { + "index": 16, + "filename": "shard_00016.bin", + "size": 67108864, + "hash": "977bec777ed4fa5570b9b0726fe2212185e3eb0059c18211541fc48a321b0eb9", + "blake3": "977bec777ed4fa5570b9b0726fe2212185e3eb0059c18211541fc48a321b0eb9", + "offset": 1073741824 + }, + { + "index": 17, + "filename": "shard_00017.bin", + "size": 67108864, + "hash": "83b9856036b21ebe35fde94bef83e3ff0881bd98f7a9887faf2e4bbb1e6932d3", + "blake3": "83b9856036b21ebe35fde94bef83e3ff0881bd98f7a9887faf2e4bbb1e6932d3", + "offset": 1140850688 + }, + { + "index": 18, + "filename": "shard_00018.bin", + "size": 67108864, + "hash": "a6a8bf3a05816bd787b5b06a9adb76e65b3683cf5e41652d7d7de27a9de50202", + "blake3": "a6a8bf3a05816bd787b5b06a9adb76e65b3683cf5e41652d7d7de27a9de50202", + "offset": 1207959552 + }, + { + "index": 19, + "filename": "shard_00019.bin", + "size": 67108864, + "hash": "0eff96a75ed30b91c705cfa81d43f59ca0f48acfb22bbe72264fd47446c95c43", + "blake3": "0eff96a75ed30b91c705cfa81d43f59ca0f48acfb22bbe72264fd47446c95c43", + "offset": 1275068416 + }, + { + "index": 20, + "filename": "shard_00020.bin", + "size": 67108864, + "hash": "ea2cac26cca3c8b328bb348e05050be144e2889c669986939feaaadd81c4e5fc", + "blake3": "ea2cac26cca3c8b328bb348e05050be144e2889c669986939feaaadd81c4e5fc", + "offset": 1342177280 + }, + { + "index": 21, + "filename": "shard_00021.bin", + "size": 67108864, + "hash": "0738669d6a9ad2c95a5e4f2d1aec2f7e2397887898cca6276bd10df835e0c86a", + "blake3": "0738669d6a9ad2c95a5e4f2d1aec2f7e2397887898cca6276bd10df835e0c86a", + "offset": 1409286144 + }, + { + "index": 22, + "filename": "shard_00022.bin", + "size": 67108864, + "hash": "bca2eaa8331caee52e3db3804efc70d9b3959313affae920750f778c0317fa78", + "blake3": "bca2eaa8331caee52e3db3804efc70d9b3959313affae920750f778c0317fa78", + "offset": 1476395008 + }, + { + "index": 23, + "filename": "shard_00023.bin", + "size": 67108864, + "hash": "a38e3a0c30aa03d7971046609d0e342a74d0e7b0fd283b640dcae971415fcaca", + "blake3": "a38e3a0c30aa03d7971046609d0e342a74d0e7b0fd283b640dcae971415fcaca", + "offset": 1543503872 + }, + { + "index": 24, + "filename": "shard_00024.bin", + "size": 67108864, + "hash": "27419ab72af870d2bfed2329d8767fbd087b60720c18389ad67671c454b1991e", + "blake3": "27419ab72af870d2bfed2329d8767fbd087b60720c18389ad67671c454b1991e", + "offset": 1610612736 + }, + { + "index": 25, + "filename": "shard_00025.bin", + "size": 67108864, + "hash": "a2cbe7c0efd175eb7b585ee814111331ee6b3129b8c924f015a3cb25e06fdb0c", + "blake3": "a2cbe7c0efd175eb7b585ee814111331ee6b3129b8c924f015a3cb25e06fdb0c", + "offset": 1677721600 + }, + { + "index": 26, + "filename": "shard_00026.bin", + "size": 67108864, + "hash": "a056048fd1d2acca1c225dfcbbc030d11071064c6cec579e4968cfe08a2e4724", + "blake3": "a056048fd1d2acca1c225dfcbbc030d11071064c6cec579e4968cfe08a2e4724", + "offset": 1744830464 + }, + { + "index": 27, + "filename": "shard_00027.bin", + "size": 67108864, + "hash": "bc14ccca44b5376745601a5bc8d1d998f03eb23742d72ca0bd3e4fe6a1f75536", + "blake3": "bc14ccca44b5376745601a5bc8d1d998f03eb23742d72ca0bd3e4fe6a1f75536", + "offset": 1811939328 + }, + { + "index": 28, + "filename": "shard_00028.bin", + "size": 67108864, + "hash": "c2621431eb9086beb87d31a06c87e992ed6e5c14f63efe5f34fe5dd23f80bdb4", + "blake3": "c2621431eb9086beb87d31a06c87e992ed6e5c14f63efe5f34fe5dd23f80bdb4", + "offset": 1879048192 + }, + { + "index": 29, + "filename": "shard_00029.bin", + "size": 67108864, + "hash": "f8b4deaa15e33a52a5ad89925311e5e242a753acd7a90a1dcfd44aa63041bc39", + "blake3": "f8b4deaa15e33a52a5ad89925311e5e242a753acd7a90a1dcfd44aa63041bc39", + "offset": 1946157056 + }, + { + "index": 30, + "filename": "shard_00030.bin", + "size": 67108864, + "hash": "16dfb471ef5841cb1c73a6e7d4b9fba1d1fcb9e94a63bc5b8645589390918114", + "blake3": "16dfb471ef5841cb1c73a6e7d4b9fba1d1fcb9e94a63bc5b8645589390918114", + "offset": 2013265920 + }, + { + "index": 31, + "filename": "shard_00031.bin", + "size": 67108864, + "hash": "ebf796880de1587187329fea608d5463749dbbe22c4a8c311946fd5d87f5f417", + "blake3": "ebf796880de1587187329fea608d5463749dbbe22c4a8c311946fd5d87f5f417", + "offset": 2080374784 + }, + { + "index": 32, + "filename": "shard_00032.bin", + "size": 67108864, + "hash": "c95132af5d55cf2262346cc3f6d2f14444aba987856f69dbda85cf2aab245e6c", + "blake3": "c95132af5d55cf2262346cc3f6d2f14444aba987856f69dbda85cf2aab245e6c", + "offset": 2147483648 + }, + { + "index": 33, + "filename": "shard_00033.bin", + "size": 67108864, + "hash": "b04f7ad7206e769d8532ba69b470f851d3819339d2dfd099f159d313da9f35ca", + "blake3": "b04f7ad7206e769d8532ba69b470f851d3819339d2dfd099f159d313da9f35ca", + "offset": 2214592512 + }, + { + "index": 34, + "filename": "shard_00034.bin", + "size": 67108864, + "hash": "98b250aa99520c6912010320b0848059096377c52c409067d5bdcf7d26fc0150", + "blake3": "98b250aa99520c6912010320b0848059096377c52c409067d5bdcf7d26fc0150", + "offset": 2281701376 + }, + { + "index": 35, + "filename": "shard_00035.bin", + "size": 67108864, + "hash": "848067db11634ed5755324d0217e7703db76143f5283b2eba8e1040d60366a56", + "blake3": "848067db11634ed5755324d0217e7703db76143f5283b2eba8e1040d60366a56", + "offset": 2348810240 + }, + { + "index": 36, + "filename": "shard_00036.bin", + "size": 67108864, + "hash": "737c4c5eefc17d1b1a60837f7113cf4b238481d61c85d6fe9022629a5dbfd89c", + "blake3": "737c4c5eefc17d1b1a60837f7113cf4b238481d61c85d6fe9022629a5dbfd89c", + "offset": 2415919104 + }, + { + "index": 37, + "filename": "shard_00037.bin", + "size": 67108864, + "hash": "413ccfe177863e537fbc0bf43e94ea804642f164ff5cf61b1e8be3d2a453fa39", + "blake3": "413ccfe177863e537fbc0bf43e94ea804642f164ff5cf61b1e8be3d2a453fa39", + "offset": 2483027968 + }, + { + "index": 38, + "filename": "shard_00038.bin", + "size": 67108864, + "hash": "a6ed104eadb2606264efa0deaa7a77be7d942d1500cde7a870518477cd34031b", + "blake3": "a6ed104eadb2606264efa0deaa7a77be7d942d1500cde7a870518477cd34031b", + "offset": 2550136832 + }, + { + "index": 39, + "filename": "shard_00039.bin", + "size": 67108864, + "hash": "66bab9d832a62f161c0466fd4138591dc12060ffb909f8c60a4bce7f0b78c40c", + "blake3": "66bab9d832a62f161c0466fd4138591dc12060ffb909f8c60a4bce7f0b78c40c", + "offset": 2617245696 + }, + { + "index": 40, + "filename": "shard_00040.bin", + "size": 67108864, + "hash": "1f6b37b62b337e8ff0a20aebe749a4b7ea62b8a81870cae15cdf090681ba32b4", + "blake3": "1f6b37b62b337e8ff0a20aebe749a4b7ea62b8a81870cae15cdf090681ba32b4", + "offset": 2684354560 + }, + { + "index": 41, + "filename": "shard_00041.bin", + "size": 67108864, + "hash": "978073ea847333a6210433ee46c3b582f9c5e8bb3449ff23efdec13f12aec9b7", + "blake3": "978073ea847333a6210433ee46c3b582f9c5e8bb3449ff23efdec13f12aec9b7", + "offset": 2751463424 + }, + { + "index": 42, + "filename": "shard_00042.bin", + "size": 67108864, + "hash": "0f9a12bff371b926921789ff1529ce482e741ca524ee96b32b27a38442a83ae6", + "blake3": "0f9a12bff371b926921789ff1529ce482e741ca524ee96b32b27a38442a83ae6", + "offset": 2818572288 + }, + { + "index": 43, + "filename": "shard_00043.bin", + "size": 67108864, + "hash": "89631913602461ebf8fdec6086956ad9b5bf44052d0d11cc0e739a630848c616", + "blake3": "89631913602461ebf8fdec6086956ad9b5bf44052d0d11cc0e739a630848c616", + "offset": 2885681152 + }, + { + "index": 44, + "filename": "shard_00044.bin", + "size": 67108864, + "hash": "e8d0e59c2c805170f40f23a5c9d1450c1898856fcad4112887b01517fb5bfb2e", + "blake3": "e8d0e59c2c805170f40f23a5c9d1450c1898856fcad4112887b01517fb5bfb2e", + "offset": 2952790016 + }, + { + "index": 45, + "filename": "shard_00045.bin", + "size": 67108864, + "hash": "ceb02408794f8f53d6d18374a93ae2aaeda6687fd31dc98b21e7d66268d651d8", + "blake3": "ceb02408794f8f53d6d18374a93ae2aaeda6687fd31dc98b21e7d66268d651d8", + "offset": 3019898880 + }, + { + "index": 46, + "filename": "shard_00046.bin", + "size": 61094912, + "hash": "a7a4df74def518452779cc06d0c6a0486fa9ed6f9760a30e8e641b146d8a1bef", + "blake3": "a7a4df74def518452779cc06d0c6a0486fa9ed6f9760a30e8e641b146d8a1bef", + "offset": 3087007744 + } + ], + "tensors": { + "language_model.model.embed_tokens.weight": { + "spans": [ + { + "shardIndex": 0, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 1, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 2, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 3, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 4, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 5, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 6, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 7, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 8, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 9, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 10, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 11, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 12, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 13, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 14, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 15, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 16, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 17, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 18, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 19, + "offset": 0, + "size": 67108864 + }, + { + "shardIndex": 20, + "offset": 0, + "size": 327680 + } + ], + "size": 1342504960, + "shape": [ + 262208, + 2560 + ], + "dtype": "F16", + "role": "embedding" + }, + "language_model.model.layers.0.input_layernorm.weight": { + "shard": 20, + "offset": 327680, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.0.mlp.down_proj.weight": { + "shard": 20, + "offset": 332800, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.0.mlp.gate_proj.weight": { + "shard": 20, + "offset": 15078400, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.0.mlp.up_proj.weight": { + "shard": 20, + "offset": 29824000, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.0.post_attention_layernorm.weight": { + "shard": 20, + "offset": 44569600, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.0.post_feedforward_layernorm.weight": { + "shard": 20, + "offset": 44574720, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.0.pre_feedforward_layernorm.weight": { + "shard": 20, + "offset": 44579840, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.0.self_attn.k_norm.weight": { + "shard": 20, + "offset": 44584960, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.0.self_attn.k_proj.weight": { + "shard": 20, + "offset": 44585472, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.0.self_attn.o_proj.weight": { + "shard": 20, + "offset": 46060032, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.0.self_attn.q_norm.weight": { + "shard": 20, + "offset": 49009152, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.0.self_attn.q_proj.weight": { + "shard": 20, + "offset": 49009664, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.0.self_attn.v_proj.weight": { + "shard": 20, + "offset": 51958784, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.1.input_layernorm.weight": { + "shard": 20, + "offset": 53433344, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.1.mlp.down_proj.weight": { + "spans": [ + { + "shardIndex": 20, + "offset": 53438464, + "size": 13670400 + }, + { + "shardIndex": 21, + "offset": 0, + "size": 1075200 + } + ], + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.1.mlp.gate_proj.weight": { + "shard": 21, + "offset": 1075200, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.1.mlp.up_proj.weight": { + "shard": 21, + "offset": 15820800, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.1.post_attention_layernorm.weight": { + "shard": 21, + "offset": 30566400, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.1.post_feedforward_layernorm.weight": { + "shard": 21, + "offset": 30571520, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.1.pre_feedforward_layernorm.weight": { + "shard": 21, + "offset": 30576640, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.1.self_attn.k_norm.weight": { + "shard": 21, + "offset": 30581760, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.1.self_attn.k_proj.weight": { + "shard": 21, + "offset": 30582272, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.1.self_attn.o_proj.weight": { + "shard": 21, + "offset": 32056832, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.1.self_attn.q_norm.weight": { + "shard": 21, + "offset": 35005952, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.1.self_attn.q_proj.weight": { + "shard": 21, + "offset": 35006464, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.1.self_attn.v_proj.weight": { + "shard": 21, + "offset": 37955584, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.10.input_layernorm.weight": { + "shard": 21, + "offset": 39430144, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.10.mlp.down_proj.weight": { + "shard": 21, + "offset": 39435264, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.10.mlp.gate_proj.weight": { + "spans": [ + { + "shardIndex": 21, + "offset": 54180864, + "size": 12928000 + }, + { + "shardIndex": 22, + "offset": 0, + "size": 1817600 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.10.mlp.up_proj.weight": { + "shard": 22, + "offset": 1817600, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.10.post_attention_layernorm.weight": { + "shard": 22, + "offset": 16563200, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.10.post_feedforward_layernorm.weight": { + "shard": 22, + "offset": 16568320, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.10.pre_feedforward_layernorm.weight": { + "shard": 22, + "offset": 16573440, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.10.self_attn.k_norm.weight": { + "shard": 22, + "offset": 16578560, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.10.self_attn.k_proj.weight": { + "shard": 22, + "offset": 16579072, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.10.self_attn.o_proj.weight": { + "shard": 22, + "offset": 18053632, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.10.self_attn.q_norm.weight": { + "shard": 22, + "offset": 21002752, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.10.self_attn.q_proj.weight": { + "shard": 22, + "offset": 21003264, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.10.self_attn.v_proj.weight": { + "shard": 22, + "offset": 23952384, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.11.input_layernorm.weight": { + "shard": 22, + "offset": 25426944, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.11.mlp.down_proj.weight": { + "shard": 22, + "offset": 25432064, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.11.mlp.gate_proj.weight": { + "shard": 22, + "offset": 40177664, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.11.mlp.up_proj.weight": { + "spans": [ + { + "shardIndex": 22, + "offset": 54923264, + "size": 12185600 + }, + { + "shardIndex": 23, + "offset": 0, + "size": 2560000 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.11.post_attention_layernorm.weight": { + "shard": 23, + "offset": 2560000, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.11.post_feedforward_layernorm.weight": { + "shard": 23, + "offset": 2565120, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.11.pre_feedforward_layernorm.weight": { + "shard": 23, + "offset": 2570240, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.11.self_attn.k_norm.weight": { + "shard": 23, + "offset": 2575360, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.11.self_attn.k_proj.weight": { + "shard": 23, + "offset": 2575872, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.11.self_attn.o_proj.weight": { + "shard": 23, + "offset": 4050432, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.11.self_attn.q_norm.weight": { + "shard": 23, + "offset": 6999552, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.11.self_attn.q_proj.weight": { + "shard": 23, + "offset": 7000064, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.11.self_attn.v_proj.weight": { + "shard": 23, + "offset": 9949184, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.12.input_layernorm.weight": { + "shard": 23, + "offset": 11423744, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.12.mlp.down_proj.weight": { + "shard": 23, + "offset": 11428864, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.12.mlp.gate_proj.weight": { + "shard": 23, + "offset": 26174464, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.12.mlp.up_proj.weight": { + "shard": 23, + "offset": 40920064, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.12.post_attention_layernorm.weight": { + "shard": 23, + "offset": 55665664, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.12.post_feedforward_layernorm.weight": { + "shard": 23, + "offset": 55670784, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.12.pre_feedforward_layernorm.weight": { + "shard": 23, + "offset": 55675904, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.12.self_attn.k_norm.weight": { + "shard": 23, + "offset": 55681024, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.12.self_attn.k_proj.weight": { + "shard": 23, + "offset": 55681536, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.12.self_attn.o_proj.weight": { + "shard": 23, + "offset": 57156096, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.12.self_attn.q_norm.weight": { + "shard": 23, + "offset": 60105216, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.12.self_attn.q_proj.weight": { + "shard": 23, + "offset": 60105728, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.12.self_attn.v_proj.weight": { + "shard": 23, + "offset": 63054848, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.13.input_layernorm.weight": { + "shard": 23, + "offset": 64529408, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.13.mlp.down_proj.weight": { + "spans": [ + { + "shardIndex": 23, + "offset": 64534528, + "size": 2574336 + }, + { + "shardIndex": 24, + "offset": 0, + "size": 12171264 + } + ], + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.13.mlp.gate_proj.weight": { + "shard": 24, + "offset": 12171264, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.13.mlp.up_proj.weight": { + "shard": 24, + "offset": 26916864, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.13.post_attention_layernorm.weight": { + "shard": 24, + "offset": 41662464, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.13.post_feedforward_layernorm.weight": { + "shard": 24, + "offset": 41667584, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.13.pre_feedforward_layernorm.weight": { + "shard": 24, + "offset": 41672704, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.13.self_attn.k_norm.weight": { + "shard": 24, + "offset": 41677824, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.13.self_attn.k_proj.weight": { + "shard": 24, + "offset": 41678336, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.13.self_attn.o_proj.weight": { + "shard": 24, + "offset": 43152896, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.13.self_attn.q_norm.weight": { + "shard": 24, + "offset": 46102016, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.13.self_attn.q_proj.weight": { + "shard": 24, + "offset": 46102528, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.13.self_attn.v_proj.weight": { + "shard": 24, + "offset": 49051648, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.14.input_layernorm.weight": { + "shard": 24, + "offset": 50526208, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.14.mlp.down_proj.weight": { + "shard": 24, + "offset": 50531328, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.14.mlp.gate_proj.weight": { + "spans": [ + { + "shardIndex": 24, + "offset": 65276928, + "size": 1831936 + }, + { + "shardIndex": 25, + "offset": 0, + "size": 12913664 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.14.mlp.up_proj.weight": { + "shard": 25, + "offset": 12913664, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.14.post_attention_layernorm.weight": { + "shard": 25, + "offset": 27659264, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.14.post_feedforward_layernorm.weight": { + "shard": 25, + "offset": 27664384, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.14.pre_feedforward_layernorm.weight": { + "shard": 25, + "offset": 27669504, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.14.self_attn.k_norm.weight": { + "shard": 25, + "offset": 27674624, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.14.self_attn.k_proj.weight": { + "shard": 25, + "offset": 27675136, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.14.self_attn.o_proj.weight": { + "shard": 25, + "offset": 29149696, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.14.self_attn.q_norm.weight": { + "shard": 25, + "offset": 32098816, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.14.self_attn.q_proj.weight": { + "shard": 25, + "offset": 32099328, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.14.self_attn.v_proj.weight": { + "shard": 25, + "offset": 35048448, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.15.input_layernorm.weight": { + "shard": 25, + "offset": 36523008, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.15.mlp.down_proj.weight": { + "shard": 25, + "offset": 36528128, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.15.mlp.gate_proj.weight": { + "shard": 25, + "offset": 51273728, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.15.mlp.up_proj.weight": { + "spans": [ + { + "shardIndex": 25, + "offset": 66019328, + "size": 1089536 + }, + { + "shardIndex": 26, + "offset": 0, + "size": 13656064 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.15.post_attention_layernorm.weight": { + "shard": 26, + "offset": 13656064, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.15.post_feedforward_layernorm.weight": { + "shard": 26, + "offset": 13661184, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.15.pre_feedforward_layernorm.weight": { + "shard": 26, + "offset": 13666304, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.15.self_attn.k_norm.weight": { + "shard": 26, + "offset": 13671424, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.15.self_attn.k_proj.weight": { + "shard": 26, + "offset": 13671936, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.15.self_attn.o_proj.weight": { + "shard": 26, + "offset": 15146496, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.15.self_attn.q_norm.weight": { + "shard": 26, + "offset": 18095616, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.15.self_attn.q_proj.weight": { + "shard": 26, + "offset": 18096128, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.15.self_attn.v_proj.weight": { + "shard": 26, + "offset": 21045248, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.16.input_layernorm.weight": { + "shard": 26, + "offset": 22519808, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.16.mlp.down_proj.weight": { + "shard": 26, + "offset": 22524928, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.16.mlp.gate_proj.weight": { + "shard": 26, + "offset": 37270528, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.16.mlp.up_proj.weight": { + "shard": 26, + "offset": 52016128, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.16.post_attention_layernorm.weight": { + "shard": 26, + "offset": 66761728, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.16.post_feedforward_layernorm.weight": { + "shard": 26, + "offset": 66766848, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.16.pre_feedforward_layernorm.weight": { + "shard": 26, + "offset": 66771968, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.16.self_attn.k_norm.weight": { + "shard": 26, + "offset": 66777088, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.16.self_attn.k_proj.weight": { + "spans": [ + { + "shardIndex": 26, + "offset": 66777600, + "size": 331264 + }, + { + "shardIndex": 27, + "offset": 0, + "size": 1143296 + } + ], + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.16.self_attn.o_proj.weight": { + "shard": 27, + "offset": 1143296, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.16.self_attn.q_norm.weight": { + "shard": 27, + "offset": 4092416, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.16.self_attn.q_proj.weight": { + "shard": 27, + "offset": 4092928, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.16.self_attn.v_proj.weight": { + "shard": 27, + "offset": 7042048, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.17.input_layernorm.weight": { + "shard": 27, + "offset": 8516608, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.17.mlp.down_proj.weight": { + "shard": 27, + "offset": 8521728, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.17.mlp.gate_proj.weight": { + "shard": 27, + "offset": 23267328, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.17.mlp.up_proj.weight": { + "shard": 27, + "offset": 38012928, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.17.post_attention_layernorm.weight": { + "shard": 27, + "offset": 52758528, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.17.post_feedforward_layernorm.weight": { + "shard": 27, + "offset": 52763648, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.17.pre_feedforward_layernorm.weight": { + "shard": 27, + "offset": 52768768, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.17.self_attn.k_norm.weight": { + "shard": 27, + "offset": 52773888, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.17.self_attn.k_proj.weight": { + "shard": 27, + "offset": 52774400, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.17.self_attn.o_proj.weight": { + "shard": 27, + "offset": 54248960, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.17.self_attn.q_norm.weight": { + "shard": 27, + "offset": 57198080, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.17.self_attn.q_proj.weight": { + "shard": 27, + "offset": 57198592, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.17.self_attn.v_proj.weight": { + "shard": 27, + "offset": 60147712, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.18.input_layernorm.weight": { + "shard": 27, + "offset": 61622272, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.18.mlp.down_proj.weight": { + "spans": [ + { + "shardIndex": 27, + "offset": 61627392, + "size": 5481472 + }, + { + "shardIndex": 28, + "offset": 0, + "size": 9264128 + } + ], + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.18.mlp.gate_proj.weight": { + "shard": 28, + "offset": 9264128, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.18.mlp.up_proj.weight": { + "shard": 28, + "offset": 24009728, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.18.post_attention_layernorm.weight": { + "shard": 28, + "offset": 38755328, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.18.post_feedforward_layernorm.weight": { + "shard": 28, + "offset": 38760448, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.18.pre_feedforward_layernorm.weight": { + "shard": 28, + "offset": 38765568, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.18.self_attn.k_norm.weight": { + "shard": 28, + "offset": 38770688, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.18.self_attn.k_proj.weight": { + "shard": 28, + "offset": 38771200, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.18.self_attn.o_proj.weight": { + "shard": 28, + "offset": 40245760, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.18.self_attn.q_norm.weight": { + "shard": 28, + "offset": 43194880, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.18.self_attn.q_proj.weight": { + "shard": 28, + "offset": 43195392, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.18.self_attn.v_proj.weight": { + "shard": 28, + "offset": 46144512, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.19.input_layernorm.weight": { + "shard": 28, + "offset": 47619072, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.19.mlp.down_proj.weight": { + "shard": 28, + "offset": 47624192, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.19.mlp.gate_proj.weight": { + "spans": [ + { + "shardIndex": 28, + "offset": 62369792, + "size": 4739072 + }, + { + "shardIndex": 29, + "offset": 0, + "size": 10006528 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.19.mlp.up_proj.weight": { + "shard": 29, + "offset": 10006528, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.19.post_attention_layernorm.weight": { + "shard": 29, + "offset": 24752128, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.19.post_feedforward_layernorm.weight": { + "shard": 29, + "offset": 24757248, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.19.pre_feedforward_layernorm.weight": { + "shard": 29, + "offset": 24762368, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.19.self_attn.k_norm.weight": { + "shard": 29, + "offset": 24767488, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.19.self_attn.k_proj.weight": { + "shard": 29, + "offset": 24768000, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.19.self_attn.o_proj.weight": { + "shard": 29, + "offset": 26242560, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.19.self_attn.q_norm.weight": { + "shard": 29, + "offset": 29191680, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.19.self_attn.q_proj.weight": { + "shard": 29, + "offset": 29192192, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.19.self_attn.v_proj.weight": { + "shard": 29, + "offset": 32141312, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.2.input_layernorm.weight": { + "shard": 29, + "offset": 33615872, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.2.mlp.down_proj.weight": { + "shard": 29, + "offset": 33620992, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.2.mlp.gate_proj.weight": { + "shard": 29, + "offset": 48366592, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.2.mlp.up_proj.weight": { + "spans": [ + { + "shardIndex": 29, + "offset": 63112192, + "size": 3996672 + }, + { + "shardIndex": 30, + "offset": 0, + "size": 10748928 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.2.post_attention_layernorm.weight": { + "shard": 30, + "offset": 10748928, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.2.post_feedforward_layernorm.weight": { + "shard": 30, + "offset": 10754048, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.2.pre_feedforward_layernorm.weight": { + "shard": 30, + "offset": 10759168, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.2.self_attn.k_norm.weight": { + "shard": 30, + "offset": 10764288, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.2.self_attn.k_proj.weight": { + "shard": 30, + "offset": 10764800, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.2.self_attn.o_proj.weight": { + "shard": 30, + "offset": 12239360, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.2.self_attn.q_norm.weight": { + "shard": 30, + "offset": 15188480, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.2.self_attn.q_proj.weight": { + "shard": 30, + "offset": 15188992, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.2.self_attn.v_proj.weight": { + "shard": 30, + "offset": 18138112, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.20.input_layernorm.weight": { + "shard": 30, + "offset": 19612672, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.20.mlp.down_proj.weight": { + "shard": 30, + "offset": 19617792, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.20.mlp.gate_proj.weight": { + "shard": 30, + "offset": 34363392, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.20.mlp.up_proj.weight": { + "shard": 30, + "offset": 49108992, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.20.post_attention_layernorm.weight": { + "shard": 30, + "offset": 63854592, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.20.post_feedforward_layernorm.weight": { + "shard": 30, + "offset": 63859712, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.20.pre_feedforward_layernorm.weight": { + "shard": 30, + "offset": 63864832, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.20.self_attn.k_norm.weight": { + "shard": 30, + "offset": 63869952, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.20.self_attn.k_proj.weight": { + "shard": 30, + "offset": 63870464, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.20.self_attn.o_proj.weight": { + "spans": [ + { + "shardIndex": 30, + "offset": 65345024, + "size": 1763840 + }, + { + "shardIndex": 31, + "offset": 0, + "size": 1185280 + } + ], + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.20.self_attn.q_norm.weight": { + "shard": 31, + "offset": 1185280, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.20.self_attn.q_proj.weight": { + "shard": 31, + "offset": 1185792, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.20.self_attn.v_proj.weight": { + "shard": 31, + "offset": 4134912, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.21.input_layernorm.weight": { + "shard": 31, + "offset": 5609472, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.21.mlp.down_proj.weight": { + "shard": 31, + "offset": 5614592, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.21.mlp.gate_proj.weight": { + "shard": 31, + "offset": 20360192, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.21.mlp.up_proj.weight": { + "shard": 31, + "offset": 35105792, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.21.post_attention_layernorm.weight": { + "shard": 31, + "offset": 49851392, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.21.post_feedforward_layernorm.weight": { + "shard": 31, + "offset": 49856512, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.21.pre_feedforward_layernorm.weight": { + "shard": 31, + "offset": 49861632, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.21.self_attn.k_norm.weight": { + "shard": 31, + "offset": 49866752, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.21.self_attn.k_proj.weight": { + "shard": 31, + "offset": 49867264, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.21.self_attn.o_proj.weight": { + "shard": 31, + "offset": 51341824, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.21.self_attn.q_norm.weight": { + "shard": 31, + "offset": 54290944, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.21.self_attn.q_proj.weight": { + "shard": 31, + "offset": 54291456, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.21.self_attn.v_proj.weight": { + "shard": 31, + "offset": 57240576, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.22.input_layernorm.weight": { + "shard": 31, + "offset": 58715136, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.22.mlp.down_proj.weight": { + "spans": [ + { + "shardIndex": 31, + "offset": 58720256, + "size": 8388608 + }, + { + "shardIndex": 32, + "offset": 0, + "size": 6356992 + } + ], + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.22.mlp.gate_proj.weight": { + "shard": 32, + "offset": 6356992, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.22.mlp.up_proj.weight": { + "shard": 32, + "offset": 21102592, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.22.post_attention_layernorm.weight": { + "shard": 32, + "offset": 35848192, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.22.post_feedforward_layernorm.weight": { + "shard": 32, + "offset": 35853312, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.22.pre_feedforward_layernorm.weight": { + "shard": 32, + "offset": 35858432, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.22.self_attn.k_norm.weight": { + "shard": 32, + "offset": 35863552, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.22.self_attn.k_proj.weight": { + "shard": 32, + "offset": 35864064, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.22.self_attn.o_proj.weight": { + "shard": 32, + "offset": 37338624, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.22.self_attn.q_norm.weight": { + "shard": 32, + "offset": 40287744, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.22.self_attn.q_proj.weight": { + "shard": 32, + "offset": 40288256, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.22.self_attn.v_proj.weight": { + "shard": 32, + "offset": 43237376, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.23.input_layernorm.weight": { + "shard": 32, + "offset": 44711936, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.23.mlp.down_proj.weight": { + "shard": 32, + "offset": 44717056, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.23.mlp.gate_proj.weight": { + "spans": [ + { + "shardIndex": 32, + "offset": 59462656, + "size": 7646208 + }, + { + "shardIndex": 33, + "offset": 0, + "size": 7099392 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.23.mlp.up_proj.weight": { + "shard": 33, + "offset": 7099392, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.23.post_attention_layernorm.weight": { + "shard": 33, + "offset": 21844992, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.23.post_feedforward_layernorm.weight": { + "shard": 33, + "offset": 21850112, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.23.pre_feedforward_layernorm.weight": { + "shard": 33, + "offset": 21855232, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.23.self_attn.k_norm.weight": { + "shard": 33, + "offset": 21860352, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.23.self_attn.k_proj.weight": { + "shard": 33, + "offset": 21860864, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.23.self_attn.o_proj.weight": { + "shard": 33, + "offset": 23335424, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.23.self_attn.q_norm.weight": { + "shard": 33, + "offset": 26284544, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.23.self_attn.q_proj.weight": { + "shard": 33, + "offset": 26285056, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.23.self_attn.v_proj.weight": { + "shard": 33, + "offset": 29234176, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.24.input_layernorm.weight": { + "shard": 33, + "offset": 30708736, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.24.mlp.down_proj.weight": { + "shard": 33, + "offset": 30713856, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.24.mlp.gate_proj.weight": { + "shard": 33, + "offset": 45459456, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.24.mlp.up_proj.weight": { + "spans": [ + { + "shardIndex": 33, + "offset": 60205056, + "size": 6903808 + }, + { + "shardIndex": 34, + "offset": 0, + "size": 7841792 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.24.post_attention_layernorm.weight": { + "shard": 34, + "offset": 7841792, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.24.post_feedforward_layernorm.weight": { + "shard": 34, + "offset": 7846912, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.24.pre_feedforward_layernorm.weight": { + "shard": 34, + "offset": 7852032, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.24.self_attn.k_norm.weight": { + "shard": 34, + "offset": 7857152, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.24.self_attn.k_proj.weight": { + "shard": 34, + "offset": 7857664, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.24.self_attn.o_proj.weight": { + "shard": 34, + "offset": 9332224, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.24.self_attn.q_norm.weight": { + "shard": 34, + "offset": 12281344, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.24.self_attn.q_proj.weight": { + "shard": 34, + "offset": 12281856, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.24.self_attn.v_proj.weight": { + "shard": 34, + "offset": 15230976, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.25.input_layernorm.weight": { + "shard": 34, + "offset": 16705536, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.25.mlp.down_proj.weight": { + "shard": 34, + "offset": 16710656, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.25.mlp.gate_proj.weight": { + "shard": 34, + "offset": 31456256, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.25.mlp.up_proj.weight": { + "shard": 34, + "offset": 46201856, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.25.post_attention_layernorm.weight": { + "shard": 34, + "offset": 60947456, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.25.post_feedforward_layernorm.weight": { + "shard": 34, + "offset": 60952576, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.25.pre_feedforward_layernorm.weight": { + "shard": 34, + "offset": 60957696, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.25.self_attn.k_norm.weight": { + "shard": 34, + "offset": 60962816, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.25.self_attn.k_proj.weight": { + "shard": 34, + "offset": 60963328, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.25.self_attn.o_proj.weight": { + "shard": 34, + "offset": 62437888, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.25.self_attn.q_norm.weight": { + "shard": 34, + "offset": 65387008, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.25.self_attn.q_proj.weight": { + "spans": [ + { + "shardIndex": 34, + "offset": 65387520, + "size": 1721344 + }, + { + "shardIndex": 35, + "offset": 0, + "size": 1227776 + } + ], + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.25.self_attn.v_proj.weight": { + "shard": 35, + "offset": 1227776, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.26.input_layernorm.weight": { + "shard": 35, + "offset": 2702336, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.26.mlp.down_proj.weight": { + "shard": 35, + "offset": 2707456, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.26.mlp.gate_proj.weight": { + "shard": 35, + "offset": 17453056, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.26.mlp.up_proj.weight": { + "shard": 35, + "offset": 32198656, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.26.post_attention_layernorm.weight": { + "shard": 35, + "offset": 46944256, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.26.post_feedforward_layernorm.weight": { + "shard": 35, + "offset": 46949376, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.26.pre_feedforward_layernorm.weight": { + "shard": 35, + "offset": 46954496, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.26.self_attn.k_norm.weight": { + "shard": 35, + "offset": 46959616, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.26.self_attn.k_proj.weight": { + "shard": 35, + "offset": 46960128, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.26.self_attn.o_proj.weight": { + "shard": 35, + "offset": 48434688, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.26.self_attn.q_norm.weight": { + "shard": 35, + "offset": 51383808, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.26.self_attn.q_proj.weight": { + "shard": 35, + "offset": 51384320, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.26.self_attn.v_proj.weight": { + "shard": 35, + "offset": 54333440, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.27.input_layernorm.weight": { + "shard": 35, + "offset": 55808000, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.27.mlp.down_proj.weight": { + "spans": [ + { + "shardIndex": 35, + "offset": 55813120, + "size": 11295744 + }, + { + "shardIndex": 36, + "offset": 0, + "size": 3449856 + } + ], + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.27.mlp.gate_proj.weight": { + "shard": 36, + "offset": 3449856, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.27.mlp.up_proj.weight": { + "shard": 36, + "offset": 18195456, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.27.post_attention_layernorm.weight": { + "shard": 36, + "offset": 32941056, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.27.post_feedforward_layernorm.weight": { + "shard": 36, + "offset": 32946176, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.27.pre_feedforward_layernorm.weight": { + "shard": 36, + "offset": 32951296, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.27.self_attn.k_norm.weight": { + "shard": 36, + "offset": 32956416, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.27.self_attn.k_proj.weight": { + "shard": 36, + "offset": 32956928, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.27.self_attn.o_proj.weight": { + "shard": 36, + "offset": 34431488, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.27.self_attn.q_norm.weight": { + "shard": 36, + "offset": 37380608, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.27.self_attn.q_proj.weight": { + "shard": 36, + "offset": 37381120, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.27.self_attn.v_proj.weight": { + "shard": 36, + "offset": 40330240, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.28.input_layernorm.weight": { + "shard": 36, + "offset": 41804800, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.28.mlp.down_proj.weight": { + "shard": 36, + "offset": 41809920, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.28.mlp.gate_proj.weight": { + "spans": [ + { + "shardIndex": 36, + "offset": 56555520, + "size": 10553344 + }, + { + "shardIndex": 37, + "offset": 0, + "size": 4192256 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.28.mlp.up_proj.weight": { + "shard": 37, + "offset": 4192256, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.28.post_attention_layernorm.weight": { + "shard": 37, + "offset": 18937856, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.28.post_feedforward_layernorm.weight": { + "shard": 37, + "offset": 18942976, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.28.pre_feedforward_layernorm.weight": { + "shard": 37, + "offset": 18948096, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.28.self_attn.k_norm.weight": { + "shard": 37, + "offset": 18953216, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.28.self_attn.k_proj.weight": { + "shard": 37, + "offset": 18953728, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.28.self_attn.o_proj.weight": { + "shard": 37, + "offset": 20428288, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.28.self_attn.q_norm.weight": { + "shard": 37, + "offset": 23377408, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.28.self_attn.q_proj.weight": { + "shard": 37, + "offset": 23377920, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.28.self_attn.v_proj.weight": { + "shard": 37, + "offset": 26327040, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.29.input_layernorm.weight": { + "shard": 37, + "offset": 27801600, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.29.mlp.down_proj.weight": { + "shard": 37, + "offset": 27806720, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.29.mlp.gate_proj.weight": { + "shard": 37, + "offset": 42552320, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.29.mlp.up_proj.weight": { + "spans": [ + { + "shardIndex": 37, + "offset": 57297920, + "size": 9810944 + }, + { + "shardIndex": 38, + "offset": 0, + "size": 4934656 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.29.post_attention_layernorm.weight": { + "shard": 38, + "offset": 4934656, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.29.post_feedforward_layernorm.weight": { + "shard": 38, + "offset": 4939776, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.29.pre_feedforward_layernorm.weight": { + "shard": 38, + "offset": 4944896, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.29.self_attn.k_norm.weight": { + "shard": 38, + "offset": 4950016, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.29.self_attn.k_proj.weight": { + "shard": 38, + "offset": 4950528, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.29.self_attn.o_proj.weight": { + "shard": 38, + "offset": 6425088, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.29.self_attn.q_norm.weight": { + "shard": 38, + "offset": 9374208, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.29.self_attn.q_proj.weight": { + "shard": 38, + "offset": 9374720, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.29.self_attn.v_proj.weight": { + "shard": 38, + "offset": 12323840, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.3.input_layernorm.weight": { + "shard": 38, + "offset": 13798400, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.3.mlp.down_proj.weight": { + "shard": 38, + "offset": 13803520, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.3.mlp.gate_proj.weight": { + "shard": 38, + "offset": 28549120, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.3.mlp.up_proj.weight": { + "shard": 38, + "offset": 43294720, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.3.post_attention_layernorm.weight": { + "shard": 38, + "offset": 58040320, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.3.post_feedforward_layernorm.weight": { + "shard": 38, + "offset": 58045440, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.3.pre_feedforward_layernorm.weight": { + "shard": 38, + "offset": 58050560, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.3.self_attn.k_norm.weight": { + "shard": 38, + "offset": 58055680, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.3.self_attn.k_proj.weight": { + "shard": 38, + "offset": 58056192, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.3.self_attn.o_proj.weight": { + "shard": 38, + "offset": 59530752, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.3.self_attn.q_norm.weight": { + "shard": 38, + "offset": 62479872, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.3.self_attn.q_proj.weight": { + "shard": 38, + "offset": 62480384, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.3.self_attn.v_proj.weight": { + "shard": 38, + "offset": 65429504, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.30.input_layernorm.weight": { + "shard": 38, + "offset": 66904064, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.30.mlp.down_proj.weight": { + "spans": [ + { + "shardIndex": 38, + "offset": 66909184, + "size": 199680 + }, + { + "shardIndex": 39, + "offset": 0, + "size": 14545920 + } + ], + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.30.mlp.gate_proj.weight": { + "shard": 39, + "offset": 14545920, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.30.mlp.up_proj.weight": { + "shard": 39, + "offset": 29291520, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.30.post_attention_layernorm.weight": { + "shard": 39, + "offset": 44037120, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.30.post_feedforward_layernorm.weight": { + "shard": 39, + "offset": 44042240, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.30.pre_feedforward_layernorm.weight": { + "shard": 39, + "offset": 44047360, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.30.self_attn.k_norm.weight": { + "shard": 39, + "offset": 44052480, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.30.self_attn.k_proj.weight": { + "shard": 39, + "offset": 44052992, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.30.self_attn.o_proj.weight": { + "shard": 39, + "offset": 45527552, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.30.self_attn.q_norm.weight": { + "shard": 39, + "offset": 48476672, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.30.self_attn.q_proj.weight": { + "shard": 39, + "offset": 48477184, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.30.self_attn.v_proj.weight": { + "shard": 39, + "offset": 51426304, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.31.input_layernorm.weight": { + "shard": 39, + "offset": 52900864, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.31.mlp.down_proj.weight": { + "spans": [ + { + "shardIndex": 39, + "offset": 52905984, + "size": 14202880 + }, + { + "shardIndex": 40, + "offset": 0, + "size": 542720 + } + ], + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.31.mlp.gate_proj.weight": { + "shard": 40, + "offset": 542720, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.31.mlp.up_proj.weight": { + "shard": 40, + "offset": 15288320, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.31.post_attention_layernorm.weight": { + "shard": 40, + "offset": 30033920, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.31.post_feedforward_layernorm.weight": { + "shard": 40, + "offset": 30039040, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.31.pre_feedforward_layernorm.weight": { + "shard": 40, + "offset": 30044160, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.31.self_attn.k_norm.weight": { + "shard": 40, + "offset": 30049280, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.31.self_attn.k_proj.weight": { + "shard": 40, + "offset": 30049792, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.31.self_attn.o_proj.weight": { + "shard": 40, + "offset": 31524352, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.31.self_attn.q_norm.weight": { + "shard": 40, + "offset": 34473472, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.31.self_attn.q_proj.weight": { + "shard": 40, + "offset": 34473984, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.31.self_attn.v_proj.weight": { + "shard": 40, + "offset": 37423104, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.32.input_layernorm.weight": { + "shard": 40, + "offset": 38897664, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.32.mlp.down_proj.weight": { + "shard": 40, + "offset": 38902784, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.32.mlp.gate_proj.weight": { + "spans": [ + { + "shardIndex": 40, + "offset": 53648384, + "size": 13460480 + }, + { + "shardIndex": 41, + "offset": 0, + "size": 1285120 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.32.mlp.up_proj.weight": { + "shard": 41, + "offset": 1285120, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.32.post_attention_layernorm.weight": { + "shard": 41, + "offset": 16030720, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.32.post_feedforward_layernorm.weight": { + "shard": 41, + "offset": 16035840, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.32.pre_feedforward_layernorm.weight": { + "shard": 41, + "offset": 16040960, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.32.self_attn.k_norm.weight": { + "shard": 41, + "offset": 16046080, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.32.self_attn.k_proj.weight": { + "shard": 41, + "offset": 16046592, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.32.self_attn.o_proj.weight": { + "shard": 41, + "offset": 17521152, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.32.self_attn.q_norm.weight": { + "shard": 41, + "offset": 20470272, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.32.self_attn.q_proj.weight": { + "shard": 41, + "offset": 20470784, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.32.self_attn.v_proj.weight": { + "shard": 41, + "offset": 23419904, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.33.input_layernorm.weight": { + "shard": 41, + "offset": 24894464, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.33.mlp.down_proj.weight": { + "shard": 41, + "offset": 24899584, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.33.mlp.gate_proj.weight": { + "shard": 41, + "offset": 39645184, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.33.mlp.up_proj.weight": { + "spans": [ + { + "shardIndex": 41, + "offset": 54390784, + "size": 12718080 + }, + { + "shardIndex": 42, + "offset": 0, + "size": 2027520 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.33.post_attention_layernorm.weight": { + "shard": 42, + "offset": 2027520, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.33.post_feedforward_layernorm.weight": { + "shard": 42, + "offset": 2032640, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.33.pre_feedforward_layernorm.weight": { + "shard": 42, + "offset": 2037760, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.33.self_attn.k_norm.weight": { + "shard": 42, + "offset": 2042880, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.33.self_attn.k_proj.weight": { + "shard": 42, + "offset": 2043392, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.33.self_attn.o_proj.weight": { + "shard": 42, + "offset": 3517952, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.33.self_attn.q_norm.weight": { + "shard": 42, + "offset": 6467072, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.33.self_attn.q_proj.weight": { + "shard": 42, + "offset": 6467584, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.33.self_attn.v_proj.weight": { + "shard": 42, + "offset": 9416704, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.4.input_layernorm.weight": { + "shard": 42, + "offset": 10891264, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.4.mlp.down_proj.weight": { + "shard": 42, + "offset": 10896384, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.4.mlp.gate_proj.weight": { + "shard": 42, + "offset": 25641984, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.4.mlp.up_proj.weight": { + "shard": 42, + "offset": 40387584, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.4.post_attention_layernorm.weight": { + "shard": 42, + "offset": 55133184, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.4.post_feedforward_layernorm.weight": { + "shard": 42, + "offset": 55138304, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.4.pre_feedforward_layernorm.weight": { + "shard": 42, + "offset": 55143424, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.4.self_attn.k_norm.weight": { + "shard": 42, + "offset": 55148544, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.4.self_attn.k_proj.weight": { + "shard": 42, + "offset": 55149056, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.4.self_attn.o_proj.weight": { + "shard": 42, + "offset": 56623616, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.4.self_attn.q_norm.weight": { + "shard": 42, + "offset": 59572736, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.4.self_attn.q_proj.weight": { + "shard": 42, + "offset": 59573248, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.4.self_attn.v_proj.weight": { + "shard": 42, + "offset": 62522368, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.5.input_layernorm.weight": { + "shard": 42, + "offset": 63996928, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.5.mlp.down_proj.weight": { + "spans": [ + { + "shardIndex": 42, + "offset": 64002048, + "size": 3106816 + }, + { + "shardIndex": 43, + "offset": 0, + "size": 11638784 + } + ], + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.5.mlp.gate_proj.weight": { + "shard": 43, + "offset": 11638784, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.5.mlp.up_proj.weight": { + "shard": 43, + "offset": 26384384, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.5.post_attention_layernorm.weight": { + "shard": 43, + "offset": 41129984, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.5.post_feedforward_layernorm.weight": { + "shard": 43, + "offset": 41135104, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.5.pre_feedforward_layernorm.weight": { + "shard": 43, + "offset": 41140224, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.5.self_attn.k_norm.weight": { + "shard": 43, + "offset": 41145344, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.5.self_attn.k_proj.weight": { + "shard": 43, + "offset": 41145856, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.5.self_attn.o_proj.weight": { + "shard": 43, + "offset": 42620416, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.5.self_attn.q_norm.weight": { + "shard": 43, + "offset": 45569536, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.5.self_attn.q_proj.weight": { + "shard": 43, + "offset": 45570048, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.5.self_attn.v_proj.weight": { + "shard": 43, + "offset": 48519168, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.6.input_layernorm.weight": { + "shard": 43, + "offset": 49993728, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.6.mlp.down_proj.weight": { + "shard": 43, + "offset": 49998848, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.6.mlp.gate_proj.weight": { + "spans": [ + { + "shardIndex": 43, + "offset": 64744448, + "size": 2364416 + }, + { + "shardIndex": 44, + "offset": 0, + "size": 12381184 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.6.mlp.up_proj.weight": { + "shard": 44, + "offset": 12381184, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.6.post_attention_layernorm.weight": { + "shard": 44, + "offset": 27126784, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.6.post_feedforward_layernorm.weight": { + "shard": 44, + "offset": 27131904, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.6.pre_feedforward_layernorm.weight": { + "shard": 44, + "offset": 27137024, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.6.self_attn.k_norm.weight": { + "shard": 44, + "offset": 27142144, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.6.self_attn.k_proj.weight": { + "shard": 44, + "offset": 27142656, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.6.self_attn.o_proj.weight": { + "shard": 44, + "offset": 28617216, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.6.self_attn.q_norm.weight": { + "shard": 44, + "offset": 31566336, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.6.self_attn.q_proj.weight": { + "shard": 44, + "offset": 31566848, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.6.self_attn.v_proj.weight": { + "shard": 44, + "offset": 34515968, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.7.input_layernorm.weight": { + "shard": 44, + "offset": 35990528, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.7.mlp.down_proj.weight": { + "shard": 44, + "offset": 35995648, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.7.mlp.gate_proj.weight": { + "shard": 44, + "offset": 50741248, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.7.mlp.up_proj.weight": { + "spans": [ + { + "shardIndex": 44, + "offset": 65486848, + "size": 1622016 + }, + { + "shardIndex": 45, + "offset": 0, + "size": 13123584 + } + ], + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.7.post_attention_layernorm.weight": { + "shard": 45, + "offset": 13123584, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.7.post_feedforward_layernorm.weight": { + "shard": 45, + "offset": 13128704, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.7.pre_feedforward_layernorm.weight": { + "shard": 45, + "offset": 13133824, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.7.self_attn.k_norm.weight": { + "shard": 45, + "offset": 13138944, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.7.self_attn.k_proj.weight": { + "shard": 45, + "offset": 13139456, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.7.self_attn.o_proj.weight": { + "shard": 45, + "offset": 14614016, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.7.self_attn.q_norm.weight": { + "shard": 45, + "offset": 17563136, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.7.self_attn.q_proj.weight": { + "shard": 45, + "offset": 17563648, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.7.self_attn.v_proj.weight": { + "shard": 45, + "offset": 20512768, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.8.input_layernorm.weight": { + "shard": 45, + "offset": 21987328, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.8.mlp.down_proj.weight": { + "shard": 45, + "offset": 21992448, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.8.mlp.gate_proj.weight": { + "shard": 45, + "offset": 36738048, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.8.mlp.up_proj.weight": { + "shard": 45, + "offset": 51483648, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.8.post_attention_layernorm.weight": { + "shard": 45, + "offset": 66229248, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.8.post_feedforward_layernorm.weight": { + "shard": 45, + "offset": 66234368, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.8.pre_feedforward_layernorm.weight": { + "shard": 45, + "offset": 66239488, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.8.self_attn.k_norm.weight": { + "shard": 45, + "offset": 66244608, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.8.self_attn.k_proj.weight": { + "spans": [ + { + "shardIndex": 45, + "offset": 66245120, + "size": 863744 + }, + { + "shardIndex": 46, + "offset": 0, + "size": 610816 + } + ], + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.8.self_attn.o_proj.weight": { + "shard": 46, + "offset": 610816, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.8.self_attn.q_norm.weight": { + "shard": 46, + "offset": 3559936, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.8.self_attn.q_proj.weight": { + "shard": 46, + "offset": 3560448, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.8.self_attn.v_proj.weight": { + "shard": 46, + "offset": 6509568, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.9.input_layernorm.weight": { + "shard": 46, + "offset": 7984128, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.9.mlp.down_proj.weight": { + "shard": 46, + "offset": 7989248, + "size": 14745600, + "shape": [ + 2560, + 10240 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.9.mlp.gate_proj.weight": { + "shard": 46, + "offset": 22734848, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.9.mlp.up_proj.weight": { + "shard": 46, + "offset": 37480448, + "size": 14745600, + "shape": [ + 10240, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.9.post_attention_layernorm.weight": { + "shard": 46, + "offset": 52226048, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.9.post_feedforward_layernorm.weight": { + "shard": 46, + "offset": 52231168, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.9.pre_feedforward_layernorm.weight": { + "shard": 46, + "offset": 52236288, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.9.self_attn.k_norm.weight": { + "shard": 46, + "offset": 52241408, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.9.self_attn.k_proj.weight": { + "shard": 46, + "offset": 52241920, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.9.self_attn.o_proj.weight": { + "shard": 46, + "offset": 53716480, + "size": 2949120, + "shape": [ + 2560, + 2048 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.9.self_attn.q_norm.weight": { + "shard": 46, + "offset": 56665600, + "size": 512, + "shape": [ + 256 + ], + "dtype": "BF16", + "role": "norm" + }, + "language_model.model.layers.9.self_attn.q_proj.weight": { + "shard": 46, + "offset": 56666112, + "size": 2949120, + "shape": [ + 2048, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.layers.9.self_attn.v_proj.weight": { + "shard": 46, + "offset": 59615232, + "size": 1474560, + "shape": [ + 1024, + 2560 + ], + "dtype": "Q4_K_M", + "role": "matmul", + "layout": "row" + }, + "language_model.model.norm.weight": { + "shard": 46, + "offset": 61089792, + "size": 5120, + "shape": [ + 2560 + ], + "dtype": "BF16", + "role": "norm" + } + }, + "totalSize": 3148102656, + "hashAlgorithm": "blake3", + "eos_token_id": 1, + "metadata": { + "source": "convert-core", + "convertedAt": "2026-03-02T16:40:56.319Z", + "hasTokenizer": true, + "manifestRefresh": { + "at": "2026-03-06T01:48:16.162Z", + "config": "translategemma-4b-it-wq4k-ef16-hf16.json" + } + }, + "tokenizer": { + "type": "bundled", + "vocabSize": 3119, + "file": "tokenizer.json" + } +} \ No newline at end of file diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00000.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00000.bin new file mode 100644 index 0000000000000000000000000000000000000000..f47a088184668f0f2198896dc9b025f3e1738e0b --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00000.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cf80564359f0afeebb035382bd2dc79cfeb102c5b4ea9db543648e1842be482d +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00001.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00001.bin new file mode 100644 index 0000000000000000000000000000000000000000..b1d254654678999f0ffcdf4e724a2fd924547918 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00001.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9895a5ed10942433e2a88ba133d5a59b6df5979e4ed52fce01fd91e8f9dbddc4 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00002.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00002.bin new file mode 100644 index 0000000000000000000000000000000000000000..9c0315bad27827ac49c9e8fbe1789ff7fd26aa7f --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00002.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:992238de72fff231c214466dc7406b6db767b721d6b164ef5e873b306dc064ce +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00003.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00003.bin new file mode 100644 index 0000000000000000000000000000000000000000..5fdc3abf36313a6cd356a61e68c5fd67fe886410 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00003.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:97b19561f7684a1f49c7539e0d4652da04748e07be88e0a664194b785100c866 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00004.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00004.bin new file mode 100644 index 0000000000000000000000000000000000000000..2b8d5a32e5461a0e7b7c7bb5fb27111578447281 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00004.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e3880a6795f6f3948669a32b0e9cda132cb376845d7f35914c4948a57bb7fcc +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00005.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00005.bin new file mode 100644 index 0000000000000000000000000000000000000000..aa642d76d050c3b9f1e11d880b4a8189a633057a --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00005.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:676888001d5b5b3a9124884c984956d8c765e93ab6c3adf3a3dca53ec54bf013 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00006.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00006.bin new file mode 100644 index 0000000000000000000000000000000000000000..c6dbf67c80efcb6f2b84e9bde20bee6a41a60ab2 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00006.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90168832e7c3274c1da8540b17b77913d51c82663c020997388cf67f05e5eb37 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00007.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00007.bin new file mode 100644 index 0000000000000000000000000000000000000000..fac05086f9655241ee17b38c5388803da53671cd --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00007.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c2de57e4e575a1937c4f84e46d3ef088b64e4139063ddbb55dbabbd4b8d0fce7 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00008.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00008.bin new file mode 100644 index 0000000000000000000000000000000000000000..14a6a8fa074b7323bbaa6e9ff993ba0237213c78 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00008.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:76b4b310dc1674093ae36410d3db73fa33163bb315504562f81fbc166f7cb7cd +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00009.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00009.bin new file mode 100644 index 0000000000000000000000000000000000000000..ea0720ed32211f02cdac33984b5147f03f7af702 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00009.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:00e9736d661d0c5e7ad9c50704c40add77dd35aaebc87ec89dc553dcde3a279d +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00010.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00010.bin new file mode 100644 index 0000000000000000000000000000000000000000..9b3b26f0b92f7766b1af4741b21f6b87dd7b9b43 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00010.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eaee77a37abf1b2b94afcfaee3c34b0f6e0b0020ddbd293ad51aa9d02f79ab79 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00011.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00011.bin new file mode 100644 index 0000000000000000000000000000000000000000..bd9cf774a81817234e7e6d435f90163ec6a9c72f --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00011.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8db9f6270d959b7f0485b2c86b9f1beaedbeede3bfd4847848e832d9c7e56326 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00012.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00012.bin new file mode 100644 index 0000000000000000000000000000000000000000..17d39440416e60e5855ff6918425ccaedf65b9db --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00012.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:007d2e60f3fffb57a2fb541d3799a8c319c42e0ce8ffc0db5069d353cdf91573 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00013.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00013.bin new file mode 100644 index 0000000000000000000000000000000000000000..53445a86f2167e5dbac1d84f4d01a4ae246617cb --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00013.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f263faaa5e59cfd520c11e523a1ab4c484cecb203b4678b3117c8bd5fc529651 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00014.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00014.bin new file mode 100644 index 0000000000000000000000000000000000000000..9c9683f02df84d240455237bbb948db034997c2b --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00014.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2174525134958f20d14b976bc48c459fb950cff6aed99029aa601986c982b5dc +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00015.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00015.bin new file mode 100644 index 0000000000000000000000000000000000000000..080847ba95c68a1ad113050bf2116c3667b0ba96 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00015.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:735c2862d6830d1abc619d1eb9de670bb1625643aefab7117c224fa9d4589dcd +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00016.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00016.bin new file mode 100644 index 0000000000000000000000000000000000000000..32ec4befd9bce75ba486d3efbb39b51d676a7e99 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00016.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c9bc9e5b1975a6873951658def8776e31859e72e5b86e0b179c149925f6f3e43 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00017.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00017.bin new file mode 100644 index 0000000000000000000000000000000000000000..3181163117437028464eb7aeb8965a208c8724ba --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00017.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b16388695e04e3c794c1663a2e548ff8a50ca8a379ab6a4d389b71637e8eb59f +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00018.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00018.bin new file mode 100644 index 0000000000000000000000000000000000000000..d7bb809438cb745b6aefda9d94abca3c3cc512fd --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00018.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4e63216959efb7983072ce76bb3d20154b612421fd2a2c3b24457a61899351ea +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00019.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00019.bin new file mode 100644 index 0000000000000000000000000000000000000000..e74458afeb84fc14f3f07a0321476a08301e6513 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00019.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b26314b1b3a7c69b96e26d2ca2f77bae8301741d826d570ec4ec5d8981d803b9 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00020.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00020.bin new file mode 100644 index 0000000000000000000000000000000000000000..e4ac539902b8acfd5f5b978e023115ac9a1b913b --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00020.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c6cd523dee8cfc1cbe0fc028b405d41fd94528483b9b214d1081791cddb5cf77 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00021.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00021.bin new file mode 100644 index 0000000000000000000000000000000000000000..a17e55d80bf4424e1d3d12eb0a9af6eede8a5560 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00021.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f799996e54f239a08310359f1732f300d53e086a6e40ca57d50c1c9226a2f3f6 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00022.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00022.bin new file mode 100644 index 0000000000000000000000000000000000000000..215cd14d9b4b51ba4782a1d0d93f66c0408f983a --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00022.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b57e3b03c407f01f961edf5382595def80d6bce8ccbb911ad8ed9203778bdabc +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00023.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00023.bin new file mode 100644 index 0000000000000000000000000000000000000000..916515d01395278a95656bcf166aea74ce878671 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00023.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c87743f21279eb3ca25b18a22e61912e330f0792ebff87dadd9c3d480fe3af1d +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00024.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00024.bin new file mode 100644 index 0000000000000000000000000000000000000000..c145c6ed4f5ee554a041a2f59bda1155da1ed923 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00024.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e971c1df06c5769a997041219dfc9d617f06ea1e97a58f2e0de3394d879c53fd +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00025.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00025.bin new file mode 100644 index 0000000000000000000000000000000000000000..39063319b12aa3fedf092dfb24f1429abc76faf4 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00025.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d130e31916e79c197e8f5109c27403e424d577ad3573497cbf497fdd6b4ad643 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00026.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00026.bin new file mode 100644 index 0000000000000000000000000000000000000000..5086d619a645199d180ac33362ea45c055c926ac --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00026.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:76a89805e3fd1f596967b30dacbd827b8fe5077c630c0a0209707f896804f674 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00027.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00027.bin new file mode 100644 index 0000000000000000000000000000000000000000..6ed996e3a3313222f8080f612401ff31c8b6c623 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00027.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:663243299e2c2454ed0ac4b6116a19f04a8fd05e94bcba0f0edf9a4aa898c8e8 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00028.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00028.bin new file mode 100644 index 0000000000000000000000000000000000000000..75730c7d5d633d00d863e4e92e573bb22dfb35aa --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00028.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9c589b3d59afa0e4bc62aa253b9c461cc2ba9b7fd496deec73a7f647a174535d +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00029.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00029.bin new file mode 100644 index 0000000000000000000000000000000000000000..5267d5b96637ce50a2e8962f989a06282926c4bd --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00029.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7243679f13d965abdb579a2fd0ad5c3d56b3104c3786b340e84dae0ad6c9d852 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00030.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00030.bin new file mode 100644 index 0000000000000000000000000000000000000000..4ced424f55c7c0c411bfd7cd31389b7ada3c817d --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00030.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c4394e6ae16228e3f159117bf672427e513549ea7dd9acbea6c80685c9da7a9 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00031.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00031.bin new file mode 100644 index 0000000000000000000000000000000000000000..fcdd5c052bff5d914502c6f34a89c96ede71038f --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00031.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1d9822a2abbf9a468b3d04c357752942e8759842d7581f5bd873defa40a4f230 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00032.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00032.bin new file mode 100644 index 0000000000000000000000000000000000000000..b7751f66172375db4f9479ee2f359405931fd4af --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00032.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1d4e849734bab7a3a5bc342efd875e3918d016b0b586336ecb5718dbbfbe4684 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00033.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00033.bin new file mode 100644 index 0000000000000000000000000000000000000000..38f8eeafa9be81367bb11121300cf9da82766a2b --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00033.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5a9f1a9fcf429fa37421d2ed3ef7878b6ea0125bfe09112eae2375a046f84d03 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00034.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00034.bin new file mode 100644 index 0000000000000000000000000000000000000000..8bc8c7c1ad25a4b81bc1d88f8dfd204bab3ebf4e --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00034.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d0798066d148d407551295e563d49f9f65bf2685f58e40bd938007ec1a8e7a44 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00035.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00035.bin new file mode 100644 index 0000000000000000000000000000000000000000..53f03093d96251ee7ab42086c1cfa9dba500b547 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00035.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a4761b50df8ce9ef4ae19dfa4ae69eab90c27a566f9afbfdef5637e2c0497453 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00036.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00036.bin new file mode 100644 index 0000000000000000000000000000000000000000..4aa600e539720693079fc28ba8d794694be0620d --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00036.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0ce3e2ddc9424933218dd08e8a2644fbc268d0b9c3e7073ec5a741ad291e405f +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00037.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00037.bin new file mode 100644 index 0000000000000000000000000000000000000000..034603db6ef1734b778afccebd8f642083a26605 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00037.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:79188cba65d1f1e608245d421490a02b5ae801d7d02bfaa5a1ee4ec9f2ce7872 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00038.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00038.bin new file mode 100644 index 0000000000000000000000000000000000000000..527e3987580930a04ab2fd57fd9438f6a4f0bfbe --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00038.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cce7ab7cc2406c9eda7c7922cee8d7890896393a5a522b14a132f8a515a51227 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00039.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00039.bin new file mode 100644 index 0000000000000000000000000000000000000000..2035c14559c1c4c1a49217484ff542cd52a12f0d --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00039.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8a69cbfe11b88e29a4a39aec25d322d80e0fe29a713c5bedd1b04a80a7b8c5f8 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00040.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00040.bin new file mode 100644 index 0000000000000000000000000000000000000000..46c45d943d40d916841596787086ff860da36ed2 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00040.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:823c9930d4890045f57702e5e5340b64f614208e81ddd9339bff4c9b82d8e35e +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00041.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00041.bin new file mode 100644 index 0000000000000000000000000000000000000000..ff01110928839b070372dac2f92077889b512392 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00041.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:206ca3f6862ed0b99136856026f0513e6a14518dd5bb1c690bdd68db4749605a +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00042.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00042.bin new file mode 100644 index 0000000000000000000000000000000000000000..8171698eb1063b6582b84ba6cba34a16a7c2c9e2 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00042.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:99f048ca99653f3ad9816963c0b0407a606ccbc9a8cf4b7143d6f297d33bc3e4 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00043.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00043.bin new file mode 100644 index 0000000000000000000000000000000000000000..3659fc1150aa34b2fbf8ba553b3147fa01638b70 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00043.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ff2db3b3fdddceca1cf8a891ee566f061377d18d38efb47d39cc805003e451d3 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00044.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00044.bin new file mode 100644 index 0000000000000000000000000000000000000000..92e179e7cf4fed7fc06900482f2ab374fc710324 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00044.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f60e3f4c92fd382db7315dffd3804a7c001e0cd891e652fe9f2ce85a27a5aae1 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00045.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00045.bin new file mode 100644 index 0000000000000000000000000000000000000000..04c09c285896983bb8dc69e215c040e2689cb807 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00045.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:34e7b7502a4687b05a5b73963556edb86e67e1fec7ce4a6cd8e9c74451a07822 +size 67108864 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00046.bin b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00046.bin new file mode 100644 index 0000000000000000000000000000000000000000..81960bb1a739a7db00f93b28f0bcb4eef71afe06 --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/shard_00046.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:de9cead6baaccac9bba2773e01978befd891040f4447424788d9b03cb118a9cb +size 61094912 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/tokenizer.json b/models/translategemma-4b-it-wq4k-ef16-hf16/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..a66c5d48d6c8e9fb437ad1ddfecab8ffab88b7ef --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be6955513a2509d69d18e82180c2a477ad99358d15212fb81d957566d8186075 +size 14386247 diff --git a/models/translategemma-4b-it-wq4k-ef16-hf16/tokenizer.model b/models/translategemma-4b-it-wq4k-ef16-hf16/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..14f810a829755bae3fafd6f97096dbd2eac556bd --- /dev/null +++ b/models/translategemma-4b-it-wq4k-ef16-hf16/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c +size 4689074