Synchronizing local compiler cache.
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- .gitattributes +23 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/2da7a00f0478d50ae1e7f75f085c5b2773b5f355f427c61cf34cb6febd629d96/d2900659f0c3438aa013.json +59 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/441269935591cad8d370e512c0b93cdd2fce6247c40e5a4866d872ee5338b0de/7fde6971cc66d8965351.json +81 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/4ab8140bc7eb4a553d95855c5c2be2cf8c0fbab21b823d76183b6f51e98b6fc5/af2d330d6e4b2cf43703.json +58 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/6454afdf3e9d66c7226c13a575b718845c25e53b0699600ba2bb4f883e9d841b/9a1461765dbc24d49ac7.json +62 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/73707b485eab9008c7aba7f5dad0ce2384ac685318d5f888c12fa0d81ed90b19/8adf13bbd08271a796c8.json +134 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/7518518c7e077820070186deda960d8cc49db068cdf0ac70664098fa2b6b698c/ee9ffbe1ac4a7ad2ee51.json +64 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/920f44ce6d3e004d1ce547ae06644f7be262180644b04573153aa15d98742edc/a4bf640c57a2ab4c037e.json +65 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/9c24a2080ec26e9c74d5bd90866f74131aa4d5c975415b96f5e6600f081d5647/16b8e47a972c18b057c1.json +81 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/cf6b9a360dcf294104671106bae2adbd9fd291823bb60a351883163684073231/b52ec561f1c1fa96cdd0.json +63 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/d139acf64685f15794bb983ff6eb881bdd31304bae88b0ce1ed20a54c21f2265/646e661265ff77f78ecc.json +58 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/gemma3_text/unsloth/gemma-3-270m-it/7fde6971cc66d8965351.json +81 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/granite/hf-internal-testing/tiny-random-GraniteForCausalLM/646e661265ff77f78ecc.json +58 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/llama/llamafactory/tiny-random-Llama-3/9a1461765dbc24d49ac7.json +62 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/llama/unsloth/Llama-3.2-1B-Instruct/b52ec561f1c1fa96cdd0.json +63 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/llama4_text/tiny-random/llama-4/16b8e47a972c18b057c1.json +81 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/mixtral/dacorvo/Mixtral-tiny/af2d330d6e4b2cf43703.json +58 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/phi3/yujiepan/phi-4-tiny-random/d2900659f0c3438aa013.json +59 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/qwen2/yujiepan/qwen2.5-128k-tiny-random/ee9ffbe1ac4a7ad2ee51.json +64 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/qwen3_moe/optimum-internal-testing/tiny-random-qwen3_moe/a4bf640c57a2ab4c037e.json +65 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/smollm3/HuggingFaceTB/SmolLM3-3B/8adf13bbd08271a796c8.json +134 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_0d92bb2b1fa1b9d3a3d0+a02c3a36/compile_flags.json +1 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_0d92bb2b1fa1b9d3a3d0+a02c3a36/model.done +0 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_0d92bb2b1fa1b9d3a3d0+a02c3a36/model.hlo_module.pb +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_0d92bb2b1fa1b9d3a3d0+a02c3a36/model.neff +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_0d92bb2b1fa1b9d3a3d0+a02c3a36/wrapped_neff.hlo +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_142e7dededa7227b3e62+a02c3a36/compile_flags.json +1 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_142e7dededa7227b3e62+a02c3a36/model.done +0 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_142e7dededa7227b3e62+a02c3a36/model.hlo_module.pb +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_142e7dededa7227b3e62+a02c3a36/model.neff +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_142e7dededa7227b3e62+a02c3a36/wrapped_neff.hlo +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_15c98fdcb7e9199cea41+a02c3a36/compile_flags.json +1 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_15c98fdcb7e9199cea41+a02c3a36/model.done +0 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_15c98fdcb7e9199cea41+a02c3a36/model.hlo_module.pb +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_15c98fdcb7e9199cea41+a02c3a36/model.neff +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_15c98fdcb7e9199cea41+a02c3a36/wrapped_neff.hlo +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_2dd5e80c893e94b8c53c+677eeb9d/compile_flags.json +1 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_2dd5e80c893e94b8c53c+677eeb9d/model.done +0 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_2dd5e80c893e94b8c53c+677eeb9d/model.hlo_module.pb +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_2dd5e80c893e94b8c53c+677eeb9d/model.neff +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_36e9f4ac79a30a03d15c+a02c3a36/compile_flags.json +1 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_36e9f4ac79a30a03d15c+a02c3a36/model.done +0 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_36e9f4ac79a30a03d15c+a02c3a36/model.hlo_module.pb +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_36e9f4ac79a30a03d15c+a02c3a36/model.neff +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_36e9f4ac79a30a03d15c+a02c3a36/wrapped_neff.hlo +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_3ee7521f420173f2d5db+80d05c3f/compile_flags.json +1 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_3ee7521f420173f2d5db+80d05c3f/model.done +0 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_3ee7521f420173f2d5db+80d05c3f/model.hlo_module.pb +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_3ee7521f420173f2d5db+80d05c3f/model.neff +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_3ee7521f420173f2d5db+80d05c3f/wrapped_neff.hlo +3 -0
.gitattributes
CHANGED
|
@@ -7003,3 +7003,26 @@ neuronxcc-2.21.33363.0+82129205/MODULE_b590c13d190cdf017ae9+6170d8e1/model.neff
|
|
| 7003 |
neuronxcc-2.21.33363.0+82129205/MODULE_c9c3ee67891b8431ed41+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7004 |
neuronxcc-2.21.33363.0+82129205/MODULE_c9c3ee67891b8431ed41+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 7005 |
neuronxcc-2.21.33363.0+82129205/MODULE_565345fb157ec98e94fb+6170d8e1/model.neff filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7003 |
neuronxcc-2.21.33363.0+82129205/MODULE_c9c3ee67891b8431ed41+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7004 |
neuronxcc-2.21.33363.0+82129205/MODULE_c9c3ee67891b8431ed41+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 7005 |
neuronxcc-2.21.33363.0+82129205/MODULE_565345fb157ec98e94fb+6170d8e1/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7006 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_0d92bb2b1fa1b9d3a3d0+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7007 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_0d92bb2b1fa1b9d3a3d0+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 7008 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_142e7dededa7227b3e62+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7009 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_142e7dededa7227b3e62+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 7010 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_15c98fdcb7e9199cea41+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7011 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_15c98fdcb7e9199cea41+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 7012 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_2dd5e80c893e94b8c53c+677eeb9d/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7013 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_36e9f4ac79a30a03d15c+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7014 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_36e9f4ac79a30a03d15c+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 7015 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_3ee7521f420173f2d5db+80d05c3f/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7016 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_3ee7521f420173f2d5db+80d05c3f/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 7017 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_47c06cec950237374a0e+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7018 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_47c06cec950237374a0e+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 7019 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_49ea70b4a47674881bf5+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7020 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_49ea70b4a47674881bf5+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 7021 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_78ef14a5e66efce40c70+2dde74c7/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7022 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_78ef14a5e66efce40c70+2dde74c7/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 7023 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_bd91d5ab041debe25d4b+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7024 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_bd91d5ab041debe25d4b+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 7025 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_cf5d937b64d087a8e0ee+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7026 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_cf5d937b64d087a8e0ee+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 7027 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_fe6b8b40a8e1d622e71e+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 7028 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_fe6b8b40a8e1d622e71e+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/2da7a00f0478d50ae1e7f75f085c5b2773b5f355f427c61cf34cb6febd629d96/d2900659f0c3438aa013.json
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "yujiepan/phi-4-tiny-random",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"Phi3ForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"auto_map": {},
|
| 11 |
+
"dtype": "bfloat16",
|
| 12 |
+
"embd_pdrop": 0.0,
|
| 13 |
+
"hidden_act": "silu",
|
| 14 |
+
"hidden_size": 16,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"intermediate_size": 32,
|
| 17 |
+
"max_position_embeddings": 16384,
|
| 18 |
+
"model_type": "phi3",
|
| 19 |
+
"neuron": {
|
| 20 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 21 |
+
"batch_size": 1,
|
| 22 |
+
"capacity_factor": null,
|
| 23 |
+
"checkpoint_id": "yujiepan/phi-4-tiny-random",
|
| 24 |
+
"checkpoint_revision": "18a9a1168dc97ac6d128f811925670c275610f5a",
|
| 25 |
+
"continuous_batching": false,
|
| 26 |
+
"ep_degree": 1,
|
| 27 |
+
"fused_qkv": true,
|
| 28 |
+
"glu_mlp": true,
|
| 29 |
+
"local_ranks_size": 2,
|
| 30 |
+
"max_batch_size": 1,
|
| 31 |
+
"max_context_length": 1024,
|
| 32 |
+
"max_topk": 256,
|
| 33 |
+
"n_active_tokens": 1024,
|
| 34 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 35 |
+
"on_device_sampling": true,
|
| 36 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 37 |
+
"output_logits": false,
|
| 38 |
+
"pp_degree": 1,
|
| 39 |
+
"sequence_length": 1024,
|
| 40 |
+
"speculation_length": 0,
|
| 41 |
+
"start_rank_id": 0,
|
| 42 |
+
"target": "trn1",
|
| 43 |
+
"torch_dtype": "bfloat16",
|
| 44 |
+
"tp_degree": 2
|
| 45 |
+
},
|
| 46 |
+
"num_attention_heads": 2,
|
| 47 |
+
"num_hidden_layers": 2,
|
| 48 |
+
"num_key_value_heads": 1,
|
| 49 |
+
"original_max_position_embeddings": 16384,
|
| 50 |
+
"partial_rotary_factor": 1.0,
|
| 51 |
+
"resid_pdrop": 0.0,
|
| 52 |
+
"rms_norm_eps": 1e-05,
|
| 53 |
+
"rope_scaling": null,
|
| 54 |
+
"rope_theta": 250000,
|
| 55 |
+
"sliding_window": null,
|
| 56 |
+
"tie_word_embeddings": false,
|
| 57 |
+
"use_cache": true,
|
| 58 |
+
"vocab_size": 100352
|
| 59 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/441269935591cad8d370e512c0b93cdd2fce6247c40e5a4866d872ee5338b0de/7fde6971cc66d8965351.json
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "unsloth/gemma-3-270m-it",
|
| 4 |
+
"_sliding_window_pattern": 6,
|
| 5 |
+
"_task": "text-generation",
|
| 6 |
+
"architectures": [
|
| 7 |
+
"Gemma3ForCausalLM"
|
| 8 |
+
],
|
| 9 |
+
"attention_bias": false,
|
| 10 |
+
"attention_dropout": 0.0,
|
| 11 |
+
"attn_logit_softcapping": null,
|
| 12 |
+
"dtype": "bfloat16",
|
| 13 |
+
"final_logit_softcapping": null,
|
| 14 |
+
"head_dim": 256,
|
| 15 |
+
"hidden_activation": "gelu_pytorch_tanh",
|
| 16 |
+
"hidden_size": 640,
|
| 17 |
+
"initializer_range": 0.02,
|
| 18 |
+
"intermediate_size": 2048,
|
| 19 |
+
"layer_types": [
|
| 20 |
+
"sliding_attention",
|
| 21 |
+
"sliding_attention",
|
| 22 |
+
"sliding_attention",
|
| 23 |
+
"sliding_attention",
|
| 24 |
+
"sliding_attention",
|
| 25 |
+
"full_attention",
|
| 26 |
+
"sliding_attention",
|
| 27 |
+
"sliding_attention",
|
| 28 |
+
"sliding_attention",
|
| 29 |
+
"sliding_attention",
|
| 30 |
+
"sliding_attention",
|
| 31 |
+
"full_attention",
|
| 32 |
+
"sliding_attention",
|
| 33 |
+
"sliding_attention",
|
| 34 |
+
"sliding_attention",
|
| 35 |
+
"sliding_attention",
|
| 36 |
+
"sliding_attention",
|
| 37 |
+
"full_attention"
|
| 38 |
+
],
|
| 39 |
+
"max_position_embeddings": 32768,
|
| 40 |
+
"model_type": "gemma3_text",
|
| 41 |
+
"neuron": {
|
| 42 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 43 |
+
"batch_size": 1,
|
| 44 |
+
"capacity_factor": null,
|
| 45 |
+
"checkpoint_id": "unsloth/gemma-3-270m-it",
|
| 46 |
+
"checkpoint_revision": "23cf460f6bb16954176b3ddcc8d4f250501458a9",
|
| 47 |
+
"continuous_batching": false,
|
| 48 |
+
"ep_degree": 1,
|
| 49 |
+
"fused_qkv": true,
|
| 50 |
+
"glu_mlp": true,
|
| 51 |
+
"local_ranks_size": 2,
|
| 52 |
+
"max_batch_size": 1,
|
| 53 |
+
"max_context_length": 1024,
|
| 54 |
+
"max_topk": 256,
|
| 55 |
+
"n_active_tokens": 1024,
|
| 56 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 57 |
+
"on_device_sampling": true,
|
| 58 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 59 |
+
"output_logits": false,
|
| 60 |
+
"pp_degree": 1,
|
| 61 |
+
"sequence_length": 1024,
|
| 62 |
+
"speculation_length": 0,
|
| 63 |
+
"start_rank_id": 0,
|
| 64 |
+
"target": "trn1",
|
| 65 |
+
"torch_dtype": "bfloat16",
|
| 66 |
+
"tp_degree": 2
|
| 67 |
+
},
|
| 68 |
+
"num_attention_heads": 4,
|
| 69 |
+
"num_hidden_layers": 18,
|
| 70 |
+
"num_key_value_heads": 1,
|
| 71 |
+
"query_pre_attn_scalar": 256,
|
| 72 |
+
"rms_norm_eps": 1e-06,
|
| 73 |
+
"rope_local_base_freq": 10000.0,
|
| 74 |
+
"rope_scaling": null,
|
| 75 |
+
"rope_theta": 1000000.0,
|
| 76 |
+
"sliding_window": 512,
|
| 77 |
+
"unsloth_fixed": true,
|
| 78 |
+
"use_bidirectional_attention": false,
|
| 79 |
+
"use_cache": true,
|
| 80 |
+
"vocab_size": 262144
|
| 81 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/4ab8140bc7eb4a553d95855c5c2be2cf8c0fbab21b823d76183b6f51e98b6fc5/af2d330d6e4b2cf43703.json
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "dacorvo/Mixtral-tiny",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"MixtralForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"dtype": "float16",
|
| 10 |
+
"head_dim": 32,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 1024,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 3584,
|
| 15 |
+
"max_position_embeddings": 1024,
|
| 16 |
+
"model_type": "mixtral",
|
| 17 |
+
"neuron": {
|
| 18 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 19 |
+
"batch_size": 1,
|
| 20 |
+
"capacity_factor": null,
|
| 21 |
+
"checkpoint_id": "dacorvo/Mixtral-tiny",
|
| 22 |
+
"checkpoint_revision": "c557ba205ddff6ea911f4719e0d543d6c08356b6",
|
| 23 |
+
"continuous_batching": false,
|
| 24 |
+
"ep_degree": 1,
|
| 25 |
+
"fused_qkv": false,
|
| 26 |
+
"glu_mlp": true,
|
| 27 |
+
"local_ranks_size": 2,
|
| 28 |
+
"max_batch_size": 1,
|
| 29 |
+
"max_context_length": 1024,
|
| 30 |
+
"max_topk": 256,
|
| 31 |
+
"n_active_tokens": 1024,
|
| 32 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 33 |
+
"on_device_sampling": false,
|
| 34 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 35 |
+
"output_logits": false,
|
| 36 |
+
"pp_degree": 1,
|
| 37 |
+
"sequence_length": 1024,
|
| 38 |
+
"speculation_length": 0,
|
| 39 |
+
"start_rank_id": 0,
|
| 40 |
+
"target": "trn1",
|
| 41 |
+
"torch_dtype": "float16",
|
| 42 |
+
"tp_degree": 2
|
| 43 |
+
},
|
| 44 |
+
"num_attention_heads": 32,
|
| 45 |
+
"num_experts_per_tok": 2,
|
| 46 |
+
"num_hidden_layers": 2,
|
| 47 |
+
"num_key_value_heads": 8,
|
| 48 |
+
"num_local_experts": 8,
|
| 49 |
+
"output_router_logits": false,
|
| 50 |
+
"rms_norm_eps": 1e-05,
|
| 51 |
+
"rope_theta": 10000.0,
|
| 52 |
+
"router_aux_loss_coef": 0.001,
|
| 53 |
+
"router_jitter_noise": 0.0,
|
| 54 |
+
"sliding_window": 4096,
|
| 55 |
+
"tie_word_embeddings": false,
|
| 56 |
+
"use_cache": true,
|
| 57 |
+
"vocab_size": 32000
|
| 58 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/6454afdf3e9d66c7226c13a575b718845c25e53b0699600ba2bb4f883e9d841b/9a1461765dbc24d49ac7.json
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "llamafactory/tiny-random-Llama-3",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"LlamaForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"dtype": "float16",
|
| 11 |
+
"head_dim": 4,
|
| 12 |
+
"hidden_act": "silu",
|
| 13 |
+
"hidden_size": 16,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 64,
|
| 16 |
+
"max_position_embeddings": 131072,
|
| 17 |
+
"mlp_bias": false,
|
| 18 |
+
"model_type": "llama",
|
| 19 |
+
"neuron": {
|
| 20 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 21 |
+
"batch_size": 1,
|
| 22 |
+
"capacity_factor": null,
|
| 23 |
+
"checkpoint_id": "llamafactory/tiny-random-Llama-3",
|
| 24 |
+
"checkpoint_revision": "bf2a2e3bf199ad2ee96f02a3c00246c608db22a8",
|
| 25 |
+
"continuous_batching": false,
|
| 26 |
+
"ep_degree": 1,
|
| 27 |
+
"fused_qkv": true,
|
| 28 |
+
"glu_mlp": true,
|
| 29 |
+
"local_ranks_size": 2,
|
| 30 |
+
"max_batch_size": 1,
|
| 31 |
+
"max_context_length": 1024,
|
| 32 |
+
"max_topk": 256,
|
| 33 |
+
"n_active_tokens": 1024,
|
| 34 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 35 |
+
"on_device_sampling": true,
|
| 36 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 37 |
+
"output_logits": false,
|
| 38 |
+
"pp_degree": 1,
|
| 39 |
+
"sequence_length": 1024,
|
| 40 |
+
"speculation_length": 0,
|
| 41 |
+
"start_rank_id": 0,
|
| 42 |
+
"target": "trn1",
|
| 43 |
+
"torch_dtype": "float16",
|
| 44 |
+
"tp_degree": 2
|
| 45 |
+
},
|
| 46 |
+
"num_attention_heads": 4,
|
| 47 |
+
"num_hidden_layers": 2,
|
| 48 |
+
"num_key_value_heads": 4,
|
| 49 |
+
"pretraining_tp": 1,
|
| 50 |
+
"rms_norm_eps": 1e-05,
|
| 51 |
+
"rope_scaling": {
|
| 52 |
+
"factor": 8.0,
|
| 53 |
+
"high_freq_factor": 4.0,
|
| 54 |
+
"low_freq_factor": 1.0,
|
| 55 |
+
"original_max_position_embeddings": 8192,
|
| 56 |
+
"rope_type": "llama3"
|
| 57 |
+
},
|
| 58 |
+
"rope_theta": 500000.0,
|
| 59 |
+
"tie_word_embeddings": false,
|
| 60 |
+
"use_cache": true,
|
| 61 |
+
"vocab_size": 128256
|
| 62 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/73707b485eab9008c7aba7f5dad0ce2384ac685318d5f888c12fa0d81ed90b19/8adf13bbd08271a796c8.json
ADDED
|
@@ -0,0 +1,134 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "HuggingFaceTB/SmolLM3-3B",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"SmolLM3ForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"dtype": "bfloat16",
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 2048,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 11008,
|
| 15 |
+
"layer_types": [
|
| 16 |
+
"full_attention",
|
| 17 |
+
"full_attention",
|
| 18 |
+
"full_attention",
|
| 19 |
+
"full_attention",
|
| 20 |
+
"full_attention",
|
| 21 |
+
"full_attention",
|
| 22 |
+
"full_attention",
|
| 23 |
+
"full_attention",
|
| 24 |
+
"full_attention",
|
| 25 |
+
"full_attention",
|
| 26 |
+
"full_attention",
|
| 27 |
+
"full_attention",
|
| 28 |
+
"full_attention",
|
| 29 |
+
"full_attention",
|
| 30 |
+
"full_attention",
|
| 31 |
+
"full_attention",
|
| 32 |
+
"full_attention",
|
| 33 |
+
"full_attention",
|
| 34 |
+
"full_attention",
|
| 35 |
+
"full_attention",
|
| 36 |
+
"full_attention",
|
| 37 |
+
"full_attention",
|
| 38 |
+
"full_attention",
|
| 39 |
+
"full_attention",
|
| 40 |
+
"full_attention",
|
| 41 |
+
"full_attention",
|
| 42 |
+
"full_attention",
|
| 43 |
+
"full_attention",
|
| 44 |
+
"full_attention",
|
| 45 |
+
"full_attention",
|
| 46 |
+
"full_attention",
|
| 47 |
+
"full_attention",
|
| 48 |
+
"full_attention",
|
| 49 |
+
"full_attention",
|
| 50 |
+
"full_attention",
|
| 51 |
+
"full_attention"
|
| 52 |
+
],
|
| 53 |
+
"max_position_embeddings": 65536,
|
| 54 |
+
"max_window_layers": 28,
|
| 55 |
+
"mlp_bias": false,
|
| 56 |
+
"model_type": "smollm3",
|
| 57 |
+
"neuron": {
|
| 58 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 59 |
+
"batch_size": 1,
|
| 60 |
+
"capacity_factor": null,
|
| 61 |
+
"checkpoint_id": "HuggingFaceTB/SmolLM3-3B",
|
| 62 |
+
"checkpoint_revision": "a07cc9a04f16550a088caea529712d1d335b0ac1",
|
| 63 |
+
"continuous_batching": false,
|
| 64 |
+
"ep_degree": 1,
|
| 65 |
+
"fused_qkv": true,
|
| 66 |
+
"glu_mlp": true,
|
| 67 |
+
"local_ranks_size": 2,
|
| 68 |
+
"max_batch_size": 1,
|
| 69 |
+
"max_context_length": 1024,
|
| 70 |
+
"max_topk": 256,
|
| 71 |
+
"n_active_tokens": 1024,
|
| 72 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 73 |
+
"on_device_sampling": true,
|
| 74 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 75 |
+
"output_logits": false,
|
| 76 |
+
"pp_degree": 1,
|
| 77 |
+
"sequence_length": 1024,
|
| 78 |
+
"speculation_length": 0,
|
| 79 |
+
"start_rank_id": 0,
|
| 80 |
+
"target": "trn1",
|
| 81 |
+
"torch_dtype": "bfloat16",
|
| 82 |
+
"tp_degree": 2
|
| 83 |
+
},
|
| 84 |
+
"no_rope_layer_interval": 4,
|
| 85 |
+
"no_rope_layers": [
|
| 86 |
+
1,
|
| 87 |
+
1,
|
| 88 |
+
1,
|
| 89 |
+
0,
|
| 90 |
+
1,
|
| 91 |
+
1,
|
| 92 |
+
1,
|
| 93 |
+
0,
|
| 94 |
+
1,
|
| 95 |
+
1,
|
| 96 |
+
1,
|
| 97 |
+
0,
|
| 98 |
+
1,
|
| 99 |
+
1,
|
| 100 |
+
1,
|
| 101 |
+
0,
|
| 102 |
+
1,
|
| 103 |
+
1,
|
| 104 |
+
1,
|
| 105 |
+
0,
|
| 106 |
+
1,
|
| 107 |
+
1,
|
| 108 |
+
1,
|
| 109 |
+
0,
|
| 110 |
+
1,
|
| 111 |
+
1,
|
| 112 |
+
1,
|
| 113 |
+
0,
|
| 114 |
+
1,
|
| 115 |
+
1,
|
| 116 |
+
1,
|
| 117 |
+
0,
|
| 118 |
+
1,
|
| 119 |
+
1,
|
| 120 |
+
1,
|
| 121 |
+
0
|
| 122 |
+
],
|
| 123 |
+
"num_attention_heads": 16,
|
| 124 |
+
"num_hidden_layers": 36,
|
| 125 |
+
"num_key_value_heads": 4,
|
| 126 |
+
"pretraining_tp": 2,
|
| 127 |
+
"rms_norm_eps": 1e-06,
|
| 128 |
+
"rope_scaling": null,
|
| 129 |
+
"rope_theta": 5000000.0,
|
| 130 |
+
"sliding_window": null,
|
| 131 |
+
"use_cache": false,
|
| 132 |
+
"use_sliding_window": false,
|
| 133 |
+
"vocab_size": 128256
|
| 134 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/7518518c7e077820070186deda960d8cc49db068cdf0ac70664098fa2b6b698c/ee9ffbe1ac4a7ad2ee51.json
ADDED
|
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "yujiepan/qwen2.5-128k-tiny-random",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"Qwen2ForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"dtype": "bfloat16",
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 8,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 16,
|
| 14 |
+
"layer_types": [
|
| 15 |
+
"full_attention",
|
| 16 |
+
"full_attention"
|
| 17 |
+
],
|
| 18 |
+
"max_position_embeddings": 32768,
|
| 19 |
+
"max_window_layers": 1,
|
| 20 |
+
"model_type": "qwen2",
|
| 21 |
+
"neuron": {
|
| 22 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 23 |
+
"batch_size": 1,
|
| 24 |
+
"capacity_factor": null,
|
| 25 |
+
"checkpoint_id": "yujiepan/qwen2.5-128k-tiny-random",
|
| 26 |
+
"checkpoint_revision": "c8296d4ca3f87782876d2382fbb6481d1beb8ef0",
|
| 27 |
+
"continuous_batching": false,
|
| 28 |
+
"ep_degree": 1,
|
| 29 |
+
"fused_qkv": false,
|
| 30 |
+
"glu_mlp": true,
|
| 31 |
+
"local_ranks_size": 2,
|
| 32 |
+
"max_batch_size": 1,
|
| 33 |
+
"max_context_length": 1024,
|
| 34 |
+
"max_topk": 256,
|
| 35 |
+
"n_active_tokens": 1024,
|
| 36 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 37 |
+
"on_device_sampling": true,
|
| 38 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 39 |
+
"output_logits": false,
|
| 40 |
+
"pp_degree": 1,
|
| 41 |
+
"sequence_length": 1024,
|
| 42 |
+
"speculation_length": 0,
|
| 43 |
+
"start_rank_id": 0,
|
| 44 |
+
"target": "trn1",
|
| 45 |
+
"torch_dtype": "bfloat16",
|
| 46 |
+
"tp_degree": 2
|
| 47 |
+
},
|
| 48 |
+
"num_attention_heads": 4,
|
| 49 |
+
"num_hidden_layers": 2,
|
| 50 |
+
"num_key_value_heads": 2,
|
| 51 |
+
"rms_norm_eps": 1e-06,
|
| 52 |
+
"rope_scaling": {
|
| 53 |
+
"factor": 4.0,
|
| 54 |
+
"original_max_position_embeddings": 32768,
|
| 55 |
+
"rope_type": "yarn",
|
| 56 |
+
"type": "yarn"
|
| 57 |
+
},
|
| 58 |
+
"rope_theta": 1000000.0,
|
| 59 |
+
"sliding_window": null,
|
| 60 |
+
"tie_word_embeddings": false,
|
| 61 |
+
"use_cache": true,
|
| 62 |
+
"use_sliding_window": false,
|
| 63 |
+
"vocab_size": 152064
|
| 64 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/920f44ce6d3e004d1ce547ae06644f7be262180644b04573153aa15d98742edc/a4bf640c57a2ab4c037e.json
ADDED
|
@@ -0,0 +1,65 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "optimum-internal-testing/tiny-random-qwen3_moe",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"Qwen3MoeForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"decoder_sparse_step": 2,
|
| 11 |
+
"dtype": "float32",
|
| 12 |
+
"head_dim": 32,
|
| 13 |
+
"hidden_act": "silu",
|
| 14 |
+
"hidden_size": 64,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"intermediate_size": 128,
|
| 17 |
+
"max_position_embeddings": 40960,
|
| 18 |
+
"max_window_layers": 1,
|
| 19 |
+
"mlp_only_layers": [],
|
| 20 |
+
"model_type": "qwen3_moe",
|
| 21 |
+
"moe_intermediate_size": 128,
|
| 22 |
+
"neuron": {
|
| 23 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 24 |
+
"batch_size": 1,
|
| 25 |
+
"capacity_factor": null,
|
| 26 |
+
"checkpoint_id": "optimum-internal-testing/tiny-random-qwen3_moe",
|
| 27 |
+
"checkpoint_revision": "e0230be2839556b44b7400a233c73c74b4abb7af",
|
| 28 |
+
"continuous_batching": false,
|
| 29 |
+
"ep_degree": 1,
|
| 30 |
+
"fused_qkv": false,
|
| 31 |
+
"glu_mlp": true,
|
| 32 |
+
"local_ranks_size": 2,
|
| 33 |
+
"max_batch_size": 1,
|
| 34 |
+
"max_context_length": 1024,
|
| 35 |
+
"max_topk": 256,
|
| 36 |
+
"n_active_tokens": 1024,
|
| 37 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 38 |
+
"on_device_sampling": true,
|
| 39 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 40 |
+
"output_logits": false,
|
| 41 |
+
"pp_degree": 1,
|
| 42 |
+
"sequence_length": 1024,
|
| 43 |
+
"speculation_length": 0,
|
| 44 |
+
"start_rank_id": 0,
|
| 45 |
+
"target": "trn1",
|
| 46 |
+
"torch_dtype": "float32",
|
| 47 |
+
"tp_degree": 2
|
| 48 |
+
},
|
| 49 |
+
"norm_topk_prob": true,
|
| 50 |
+
"num_attention_heads": 2,
|
| 51 |
+
"num_experts": 8,
|
| 52 |
+
"num_experts_per_tok": 2,
|
| 53 |
+
"num_hidden_layers": 2,
|
| 54 |
+
"num_key_value_heads": 1,
|
| 55 |
+
"output_router_logits": false,
|
| 56 |
+
"rms_norm_eps": 1e-06,
|
| 57 |
+
"rope_scaling": null,
|
| 58 |
+
"rope_theta": 1000000.0,
|
| 59 |
+
"router_aux_loss_coef": 0.001,
|
| 60 |
+
"sliding_window": null,
|
| 61 |
+
"tie_word_embeddings": true,
|
| 62 |
+
"use_cache": true,
|
| 63 |
+
"use_sliding_window": false,
|
| 64 |
+
"vocab_size": 151936
|
| 65 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/9c24a2080ec26e9c74d5bd90866f74131aa4d5c975415b96f5e6600f081d5647/16b8e47a972c18b057c1.json
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "tiny-random/llama-4",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_chunk_size": 128,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"attn_scale": 0.1,
|
| 9 |
+
"attn_temperature_tuning": 4,
|
| 10 |
+
"cache_implementation": "hybrid",
|
| 11 |
+
"dtype": "bfloat16",
|
| 12 |
+
"floor_scale": 8192,
|
| 13 |
+
"for_llm_compressor": false,
|
| 14 |
+
"head_dim": 32,
|
| 15 |
+
"hidden_act": "silu",
|
| 16 |
+
"hidden_size": 32,
|
| 17 |
+
"initializer_range": 0.02,
|
| 18 |
+
"interleave_moe_layer_step": 2,
|
| 19 |
+
"intermediate_size": 64,
|
| 20 |
+
"intermediate_size_mlp": 128,
|
| 21 |
+
"layer_types": [
|
| 22 |
+
"chunked_attention",
|
| 23 |
+
"chunked_attention",
|
| 24 |
+
"chunked_attention",
|
| 25 |
+
"full_attention"
|
| 26 |
+
],
|
| 27 |
+
"max_position_embeddings": 1048576,
|
| 28 |
+
"model_type": "llama4_text",
|
| 29 |
+
"moe_layers": [
|
| 30 |
+
1,
|
| 31 |
+
3
|
| 32 |
+
],
|
| 33 |
+
"neuron": {
|
| 34 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 35 |
+
"batch_size": 1,
|
| 36 |
+
"capacity_factor": null,
|
| 37 |
+
"checkpoint_id": "tiny-random/llama-4",
|
| 38 |
+
"checkpoint_revision": "9e716f5d4d1ffe0a44a15f46f4a12b840439aba4",
|
| 39 |
+
"continuous_batching": false,
|
| 40 |
+
"ep_degree": 1,
|
| 41 |
+
"fused_qkv": false,
|
| 42 |
+
"glu_mlp": true,
|
| 43 |
+
"local_ranks_size": 2,
|
| 44 |
+
"max_batch_size": 1,
|
| 45 |
+
"max_context_length": 1024,
|
| 46 |
+
"max_topk": 256,
|
| 47 |
+
"n_active_tokens": 1024,
|
| 48 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 49 |
+
"on_device_sampling": true,
|
| 50 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 51 |
+
"output_logits": false,
|
| 52 |
+
"pp_degree": 1,
|
| 53 |
+
"sequence_length": 1024,
|
| 54 |
+
"speculation_length": 0,
|
| 55 |
+
"start_rank_id": 0,
|
| 56 |
+
"target": "trn1",
|
| 57 |
+
"torch_dtype": "bfloat16",
|
| 58 |
+
"tp_degree": 2
|
| 59 |
+
},
|
| 60 |
+
"no_rope_layers": [
|
| 61 |
+
1,
|
| 62 |
+
1,
|
| 63 |
+
1,
|
| 64 |
+
0
|
| 65 |
+
],
|
| 66 |
+
"num_attention_heads": 1,
|
| 67 |
+
"num_experts_per_tok": 1,
|
| 68 |
+
"num_hidden_layers": 4,
|
| 69 |
+
"num_key_value_heads": 1,
|
| 70 |
+
"num_local_experts": 8,
|
| 71 |
+
"output_router_logits": false,
|
| 72 |
+
"rms_norm_eps": 1e-05,
|
| 73 |
+
"rope_scaling": null,
|
| 74 |
+
"rope_theta": 500000.0,
|
| 75 |
+
"router_aux_loss_coef": 0.001,
|
| 76 |
+
"router_jitter_noise": 0.0,
|
| 77 |
+
"tie_word_embeddings": true,
|
| 78 |
+
"use_cache": true,
|
| 79 |
+
"use_qk_norm": true,
|
| 80 |
+
"vocab_size": 202048
|
| 81 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/cf6b9a360dcf294104671106bae2adbd9fd291823bb60a351883163684073231/b52ec561f1c1fa96cdd0.json
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "unsloth/Llama-3.2-1B-Instruct",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"LlamaForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"dtype": "bfloat16",
|
| 11 |
+
"head_dim": 64,
|
| 12 |
+
"hidden_act": "silu",
|
| 13 |
+
"hidden_size": 2048,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 8192,
|
| 16 |
+
"max_position_embeddings": 131072,
|
| 17 |
+
"mlp_bias": false,
|
| 18 |
+
"model_type": "llama",
|
| 19 |
+
"neuron": {
|
| 20 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 21 |
+
"batch_size": 1,
|
| 22 |
+
"capacity_factor": null,
|
| 23 |
+
"checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
|
| 24 |
+
"checkpoint_revision": null,
|
| 25 |
+
"continuous_batching": false,
|
| 26 |
+
"ep_degree": 1,
|
| 27 |
+
"fused_qkv": false,
|
| 28 |
+
"glu_mlp": true,
|
| 29 |
+
"local_ranks_size": 2,
|
| 30 |
+
"max_batch_size": 1,
|
| 31 |
+
"max_context_length": 4096,
|
| 32 |
+
"max_topk": 256,
|
| 33 |
+
"n_active_tokens": 4096,
|
| 34 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 35 |
+
"on_device_sampling": false,
|
| 36 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 37 |
+
"output_logits": false,
|
| 38 |
+
"pp_degree": 1,
|
| 39 |
+
"sequence_length": 4096,
|
| 40 |
+
"speculation_length": 5,
|
| 41 |
+
"start_rank_id": 0,
|
| 42 |
+
"target": "trn1",
|
| 43 |
+
"torch_dtype": "bfloat16",
|
| 44 |
+
"tp_degree": 2
|
| 45 |
+
},
|
| 46 |
+
"num_attention_heads": 32,
|
| 47 |
+
"num_hidden_layers": 16,
|
| 48 |
+
"num_key_value_heads": 8,
|
| 49 |
+
"pretraining_tp": 1,
|
| 50 |
+
"rms_norm_eps": 1e-05,
|
| 51 |
+
"rope_scaling": {
|
| 52 |
+
"factor": 32.0,
|
| 53 |
+
"high_freq_factor": 4.0,
|
| 54 |
+
"low_freq_factor": 1.0,
|
| 55 |
+
"original_max_position_embeddings": 8192,
|
| 56 |
+
"rope_type": "llama3"
|
| 57 |
+
},
|
| 58 |
+
"rope_theta": 500000.0,
|
| 59 |
+
"tie_word_embeddings": true,
|
| 60 |
+
"unsloth_fixed": true,
|
| 61 |
+
"use_cache": true,
|
| 62 |
+
"vocab_size": 128256
|
| 63 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/d139acf64685f15794bb983ff6eb881bdd31304bae88b0ce1ed20a54c21f2265/646e661265ff77f78ecc.json
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "hf-internal-testing/tiny-random-GraniteForCausalLM",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"GraniteForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"attention_multiplier": 1.0,
|
| 11 |
+
"dtype": "float32",
|
| 12 |
+
"embedding_multiplier": 1.0,
|
| 13 |
+
"hidden_act": "silu",
|
| 14 |
+
"hidden_size": 32,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"intermediate_size": 64,
|
| 17 |
+
"logits_scaling": 1.0,
|
| 18 |
+
"max_position_embeddings": 2048,
|
| 19 |
+
"mlp_bias": false,
|
| 20 |
+
"model_type": "granite",
|
| 21 |
+
"neuron": {
|
| 22 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 23 |
+
"batch_size": 1,
|
| 24 |
+
"capacity_factor": null,
|
| 25 |
+
"checkpoint_id": "hf-internal-testing/tiny-random-GraniteForCausalLM",
|
| 26 |
+
"checkpoint_revision": "c3074ebc0ac2fe545305f5e5f6cce2cc9b2aa0c5",
|
| 27 |
+
"continuous_batching": false,
|
| 28 |
+
"ep_degree": 1,
|
| 29 |
+
"fused_qkv": true,
|
| 30 |
+
"glu_mlp": true,
|
| 31 |
+
"local_ranks_size": 2,
|
| 32 |
+
"max_batch_size": 1,
|
| 33 |
+
"max_context_length": 1024,
|
| 34 |
+
"max_topk": 256,
|
| 35 |
+
"n_active_tokens": 1024,
|
| 36 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 37 |
+
"on_device_sampling": true,
|
| 38 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 39 |
+
"output_logits": false,
|
| 40 |
+
"pp_degree": 1,
|
| 41 |
+
"sequence_length": 1024,
|
| 42 |
+
"speculation_length": 0,
|
| 43 |
+
"start_rank_id": 0,
|
| 44 |
+
"target": "trn1",
|
| 45 |
+
"torch_dtype": "float32",
|
| 46 |
+
"tp_degree": 2
|
| 47 |
+
},
|
| 48 |
+
"num_attention_heads": 4,
|
| 49 |
+
"num_hidden_layers": 2,
|
| 50 |
+
"num_key_value_heads": 4,
|
| 51 |
+
"residual_multiplier": 1.0,
|
| 52 |
+
"rms_norm_eps": 1e-06,
|
| 53 |
+
"rope_scaling": null,
|
| 54 |
+
"rope_theta": 10000.0,
|
| 55 |
+
"tie_word_embeddings": false,
|
| 56 |
+
"use_cache": true,
|
| 57 |
+
"vocab_size": 49152
|
| 58 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/gemma3_text/unsloth/gemma-3-270m-it/7fde6971cc66d8965351.json
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "unsloth/gemma-3-270m-it",
|
| 4 |
+
"_sliding_window_pattern": 6,
|
| 5 |
+
"_task": "text-generation",
|
| 6 |
+
"architectures": [
|
| 7 |
+
"Gemma3ForCausalLM"
|
| 8 |
+
],
|
| 9 |
+
"attention_bias": false,
|
| 10 |
+
"attention_dropout": 0.0,
|
| 11 |
+
"attn_logit_softcapping": null,
|
| 12 |
+
"dtype": "bfloat16",
|
| 13 |
+
"final_logit_softcapping": null,
|
| 14 |
+
"head_dim": 256,
|
| 15 |
+
"hidden_activation": "gelu_pytorch_tanh",
|
| 16 |
+
"hidden_size": 640,
|
| 17 |
+
"initializer_range": 0.02,
|
| 18 |
+
"intermediate_size": 2048,
|
| 19 |
+
"layer_types": [
|
| 20 |
+
"sliding_attention",
|
| 21 |
+
"sliding_attention",
|
| 22 |
+
"sliding_attention",
|
| 23 |
+
"sliding_attention",
|
| 24 |
+
"sliding_attention",
|
| 25 |
+
"full_attention",
|
| 26 |
+
"sliding_attention",
|
| 27 |
+
"sliding_attention",
|
| 28 |
+
"sliding_attention",
|
| 29 |
+
"sliding_attention",
|
| 30 |
+
"sliding_attention",
|
| 31 |
+
"full_attention",
|
| 32 |
+
"sliding_attention",
|
| 33 |
+
"sliding_attention",
|
| 34 |
+
"sliding_attention",
|
| 35 |
+
"sliding_attention",
|
| 36 |
+
"sliding_attention",
|
| 37 |
+
"full_attention"
|
| 38 |
+
],
|
| 39 |
+
"max_position_embeddings": 32768,
|
| 40 |
+
"model_type": "gemma3_text",
|
| 41 |
+
"neuron": {
|
| 42 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 43 |
+
"batch_size": 1,
|
| 44 |
+
"capacity_factor": null,
|
| 45 |
+
"checkpoint_id": "unsloth/gemma-3-270m-it",
|
| 46 |
+
"checkpoint_revision": "23cf460f6bb16954176b3ddcc8d4f250501458a9",
|
| 47 |
+
"continuous_batching": false,
|
| 48 |
+
"ep_degree": 1,
|
| 49 |
+
"fused_qkv": true,
|
| 50 |
+
"glu_mlp": true,
|
| 51 |
+
"local_ranks_size": 2,
|
| 52 |
+
"max_batch_size": 1,
|
| 53 |
+
"max_context_length": 1024,
|
| 54 |
+
"max_topk": 256,
|
| 55 |
+
"n_active_tokens": 1024,
|
| 56 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 57 |
+
"on_device_sampling": true,
|
| 58 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 59 |
+
"output_logits": false,
|
| 60 |
+
"pp_degree": 1,
|
| 61 |
+
"sequence_length": 1024,
|
| 62 |
+
"speculation_length": 0,
|
| 63 |
+
"start_rank_id": 0,
|
| 64 |
+
"target": "trn1",
|
| 65 |
+
"torch_dtype": "bfloat16",
|
| 66 |
+
"tp_degree": 2
|
| 67 |
+
},
|
| 68 |
+
"num_attention_heads": 4,
|
| 69 |
+
"num_hidden_layers": 18,
|
| 70 |
+
"num_key_value_heads": 1,
|
| 71 |
+
"query_pre_attn_scalar": 256,
|
| 72 |
+
"rms_norm_eps": 1e-06,
|
| 73 |
+
"rope_local_base_freq": 10000.0,
|
| 74 |
+
"rope_scaling": null,
|
| 75 |
+
"rope_theta": 1000000.0,
|
| 76 |
+
"sliding_window": 512,
|
| 77 |
+
"unsloth_fixed": true,
|
| 78 |
+
"use_bidirectional_attention": false,
|
| 79 |
+
"use_cache": true,
|
| 80 |
+
"vocab_size": 262144
|
| 81 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/granite/hf-internal-testing/tiny-random-GraniteForCausalLM/646e661265ff77f78ecc.json
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "hf-internal-testing/tiny-random-GraniteForCausalLM",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"GraniteForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"attention_multiplier": 1.0,
|
| 11 |
+
"dtype": "float32",
|
| 12 |
+
"embedding_multiplier": 1.0,
|
| 13 |
+
"hidden_act": "silu",
|
| 14 |
+
"hidden_size": 32,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"intermediate_size": 64,
|
| 17 |
+
"logits_scaling": 1.0,
|
| 18 |
+
"max_position_embeddings": 2048,
|
| 19 |
+
"mlp_bias": false,
|
| 20 |
+
"model_type": "granite",
|
| 21 |
+
"neuron": {
|
| 22 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 23 |
+
"batch_size": 1,
|
| 24 |
+
"capacity_factor": null,
|
| 25 |
+
"checkpoint_id": "hf-internal-testing/tiny-random-GraniteForCausalLM",
|
| 26 |
+
"checkpoint_revision": "c3074ebc0ac2fe545305f5e5f6cce2cc9b2aa0c5",
|
| 27 |
+
"continuous_batching": false,
|
| 28 |
+
"ep_degree": 1,
|
| 29 |
+
"fused_qkv": true,
|
| 30 |
+
"glu_mlp": true,
|
| 31 |
+
"local_ranks_size": 2,
|
| 32 |
+
"max_batch_size": 1,
|
| 33 |
+
"max_context_length": 1024,
|
| 34 |
+
"max_topk": 256,
|
| 35 |
+
"n_active_tokens": 1024,
|
| 36 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 37 |
+
"on_device_sampling": true,
|
| 38 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 39 |
+
"output_logits": false,
|
| 40 |
+
"pp_degree": 1,
|
| 41 |
+
"sequence_length": 1024,
|
| 42 |
+
"speculation_length": 0,
|
| 43 |
+
"start_rank_id": 0,
|
| 44 |
+
"target": "trn1",
|
| 45 |
+
"torch_dtype": "float32",
|
| 46 |
+
"tp_degree": 2
|
| 47 |
+
},
|
| 48 |
+
"num_attention_heads": 4,
|
| 49 |
+
"num_hidden_layers": 2,
|
| 50 |
+
"num_key_value_heads": 4,
|
| 51 |
+
"residual_multiplier": 1.0,
|
| 52 |
+
"rms_norm_eps": 1e-06,
|
| 53 |
+
"rope_scaling": null,
|
| 54 |
+
"rope_theta": 10000.0,
|
| 55 |
+
"tie_word_embeddings": false,
|
| 56 |
+
"use_cache": true,
|
| 57 |
+
"vocab_size": 49152
|
| 58 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/llama/llamafactory/tiny-random-Llama-3/9a1461765dbc24d49ac7.json
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "llamafactory/tiny-random-Llama-3",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"LlamaForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"dtype": "float16",
|
| 11 |
+
"head_dim": 4,
|
| 12 |
+
"hidden_act": "silu",
|
| 13 |
+
"hidden_size": 16,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 64,
|
| 16 |
+
"max_position_embeddings": 131072,
|
| 17 |
+
"mlp_bias": false,
|
| 18 |
+
"model_type": "llama",
|
| 19 |
+
"neuron": {
|
| 20 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 21 |
+
"batch_size": 1,
|
| 22 |
+
"capacity_factor": null,
|
| 23 |
+
"checkpoint_id": "llamafactory/tiny-random-Llama-3",
|
| 24 |
+
"checkpoint_revision": "bf2a2e3bf199ad2ee96f02a3c00246c608db22a8",
|
| 25 |
+
"continuous_batching": false,
|
| 26 |
+
"ep_degree": 1,
|
| 27 |
+
"fused_qkv": true,
|
| 28 |
+
"glu_mlp": true,
|
| 29 |
+
"local_ranks_size": 2,
|
| 30 |
+
"max_batch_size": 1,
|
| 31 |
+
"max_context_length": 1024,
|
| 32 |
+
"max_topk": 256,
|
| 33 |
+
"n_active_tokens": 1024,
|
| 34 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 35 |
+
"on_device_sampling": true,
|
| 36 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 37 |
+
"output_logits": false,
|
| 38 |
+
"pp_degree": 1,
|
| 39 |
+
"sequence_length": 1024,
|
| 40 |
+
"speculation_length": 0,
|
| 41 |
+
"start_rank_id": 0,
|
| 42 |
+
"target": "trn1",
|
| 43 |
+
"torch_dtype": "float16",
|
| 44 |
+
"tp_degree": 2
|
| 45 |
+
},
|
| 46 |
+
"num_attention_heads": 4,
|
| 47 |
+
"num_hidden_layers": 2,
|
| 48 |
+
"num_key_value_heads": 4,
|
| 49 |
+
"pretraining_tp": 1,
|
| 50 |
+
"rms_norm_eps": 1e-05,
|
| 51 |
+
"rope_scaling": {
|
| 52 |
+
"factor": 8.0,
|
| 53 |
+
"high_freq_factor": 4.0,
|
| 54 |
+
"low_freq_factor": 1.0,
|
| 55 |
+
"original_max_position_embeddings": 8192,
|
| 56 |
+
"rope_type": "llama3"
|
| 57 |
+
},
|
| 58 |
+
"rope_theta": 500000.0,
|
| 59 |
+
"tie_word_embeddings": false,
|
| 60 |
+
"use_cache": true,
|
| 61 |
+
"vocab_size": 128256
|
| 62 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/llama/unsloth/Llama-3.2-1B-Instruct/b52ec561f1c1fa96cdd0.json
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "unsloth/Llama-3.2-1B-Instruct",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"LlamaForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"dtype": "bfloat16",
|
| 11 |
+
"head_dim": 64,
|
| 12 |
+
"hidden_act": "silu",
|
| 13 |
+
"hidden_size": 2048,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 8192,
|
| 16 |
+
"max_position_embeddings": 131072,
|
| 17 |
+
"mlp_bias": false,
|
| 18 |
+
"model_type": "llama",
|
| 19 |
+
"neuron": {
|
| 20 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 21 |
+
"batch_size": 1,
|
| 22 |
+
"capacity_factor": null,
|
| 23 |
+
"checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
|
| 24 |
+
"checkpoint_revision": null,
|
| 25 |
+
"continuous_batching": false,
|
| 26 |
+
"ep_degree": 1,
|
| 27 |
+
"fused_qkv": false,
|
| 28 |
+
"glu_mlp": true,
|
| 29 |
+
"local_ranks_size": 2,
|
| 30 |
+
"max_batch_size": 1,
|
| 31 |
+
"max_context_length": 4096,
|
| 32 |
+
"max_topk": 256,
|
| 33 |
+
"n_active_tokens": 4096,
|
| 34 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 35 |
+
"on_device_sampling": false,
|
| 36 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 37 |
+
"output_logits": false,
|
| 38 |
+
"pp_degree": 1,
|
| 39 |
+
"sequence_length": 4096,
|
| 40 |
+
"speculation_length": 5,
|
| 41 |
+
"start_rank_id": 0,
|
| 42 |
+
"target": "trn1",
|
| 43 |
+
"torch_dtype": "bfloat16",
|
| 44 |
+
"tp_degree": 2
|
| 45 |
+
},
|
| 46 |
+
"num_attention_heads": 32,
|
| 47 |
+
"num_hidden_layers": 16,
|
| 48 |
+
"num_key_value_heads": 8,
|
| 49 |
+
"pretraining_tp": 1,
|
| 50 |
+
"rms_norm_eps": 1e-05,
|
| 51 |
+
"rope_scaling": {
|
| 52 |
+
"factor": 32.0,
|
| 53 |
+
"high_freq_factor": 4.0,
|
| 54 |
+
"low_freq_factor": 1.0,
|
| 55 |
+
"original_max_position_embeddings": 8192,
|
| 56 |
+
"rope_type": "llama3"
|
| 57 |
+
},
|
| 58 |
+
"rope_theta": 500000.0,
|
| 59 |
+
"tie_word_embeddings": true,
|
| 60 |
+
"unsloth_fixed": true,
|
| 61 |
+
"use_cache": true,
|
| 62 |
+
"vocab_size": 128256
|
| 63 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/llama4_text/tiny-random/llama-4/16b8e47a972c18b057c1.json
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "tiny-random/llama-4",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_chunk_size": 128,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"attn_scale": 0.1,
|
| 9 |
+
"attn_temperature_tuning": 4,
|
| 10 |
+
"cache_implementation": "hybrid",
|
| 11 |
+
"dtype": "bfloat16",
|
| 12 |
+
"floor_scale": 8192,
|
| 13 |
+
"for_llm_compressor": false,
|
| 14 |
+
"head_dim": 32,
|
| 15 |
+
"hidden_act": "silu",
|
| 16 |
+
"hidden_size": 32,
|
| 17 |
+
"initializer_range": 0.02,
|
| 18 |
+
"interleave_moe_layer_step": 2,
|
| 19 |
+
"intermediate_size": 64,
|
| 20 |
+
"intermediate_size_mlp": 128,
|
| 21 |
+
"layer_types": [
|
| 22 |
+
"chunked_attention",
|
| 23 |
+
"chunked_attention",
|
| 24 |
+
"chunked_attention",
|
| 25 |
+
"full_attention"
|
| 26 |
+
],
|
| 27 |
+
"max_position_embeddings": 1048576,
|
| 28 |
+
"model_type": "llama4_text",
|
| 29 |
+
"moe_layers": [
|
| 30 |
+
1,
|
| 31 |
+
3
|
| 32 |
+
],
|
| 33 |
+
"neuron": {
|
| 34 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 35 |
+
"batch_size": 1,
|
| 36 |
+
"capacity_factor": null,
|
| 37 |
+
"checkpoint_id": "tiny-random/llama-4",
|
| 38 |
+
"checkpoint_revision": "9e716f5d4d1ffe0a44a15f46f4a12b840439aba4",
|
| 39 |
+
"continuous_batching": false,
|
| 40 |
+
"ep_degree": 1,
|
| 41 |
+
"fused_qkv": false,
|
| 42 |
+
"glu_mlp": true,
|
| 43 |
+
"local_ranks_size": 2,
|
| 44 |
+
"max_batch_size": 1,
|
| 45 |
+
"max_context_length": 1024,
|
| 46 |
+
"max_topk": 256,
|
| 47 |
+
"n_active_tokens": 1024,
|
| 48 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 49 |
+
"on_device_sampling": true,
|
| 50 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 51 |
+
"output_logits": false,
|
| 52 |
+
"pp_degree": 1,
|
| 53 |
+
"sequence_length": 1024,
|
| 54 |
+
"speculation_length": 0,
|
| 55 |
+
"start_rank_id": 0,
|
| 56 |
+
"target": "trn1",
|
| 57 |
+
"torch_dtype": "bfloat16",
|
| 58 |
+
"tp_degree": 2
|
| 59 |
+
},
|
| 60 |
+
"no_rope_layers": [
|
| 61 |
+
1,
|
| 62 |
+
1,
|
| 63 |
+
1,
|
| 64 |
+
0
|
| 65 |
+
],
|
| 66 |
+
"num_attention_heads": 1,
|
| 67 |
+
"num_experts_per_tok": 1,
|
| 68 |
+
"num_hidden_layers": 4,
|
| 69 |
+
"num_key_value_heads": 1,
|
| 70 |
+
"num_local_experts": 8,
|
| 71 |
+
"output_router_logits": false,
|
| 72 |
+
"rms_norm_eps": 1e-05,
|
| 73 |
+
"rope_scaling": null,
|
| 74 |
+
"rope_theta": 500000.0,
|
| 75 |
+
"router_aux_loss_coef": 0.001,
|
| 76 |
+
"router_jitter_noise": 0.0,
|
| 77 |
+
"tie_word_embeddings": true,
|
| 78 |
+
"use_cache": true,
|
| 79 |
+
"use_qk_norm": true,
|
| 80 |
+
"vocab_size": 202048
|
| 81 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/mixtral/dacorvo/Mixtral-tiny/af2d330d6e4b2cf43703.json
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "dacorvo/Mixtral-tiny",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"MixtralForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"dtype": "float16",
|
| 10 |
+
"head_dim": 32,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 1024,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 3584,
|
| 15 |
+
"max_position_embeddings": 1024,
|
| 16 |
+
"model_type": "mixtral",
|
| 17 |
+
"neuron": {
|
| 18 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 19 |
+
"batch_size": 1,
|
| 20 |
+
"capacity_factor": null,
|
| 21 |
+
"checkpoint_id": "dacorvo/Mixtral-tiny",
|
| 22 |
+
"checkpoint_revision": "c557ba205ddff6ea911f4719e0d543d6c08356b6",
|
| 23 |
+
"continuous_batching": false,
|
| 24 |
+
"ep_degree": 1,
|
| 25 |
+
"fused_qkv": false,
|
| 26 |
+
"glu_mlp": true,
|
| 27 |
+
"local_ranks_size": 2,
|
| 28 |
+
"max_batch_size": 1,
|
| 29 |
+
"max_context_length": 1024,
|
| 30 |
+
"max_topk": 256,
|
| 31 |
+
"n_active_tokens": 1024,
|
| 32 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 33 |
+
"on_device_sampling": false,
|
| 34 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 35 |
+
"output_logits": false,
|
| 36 |
+
"pp_degree": 1,
|
| 37 |
+
"sequence_length": 1024,
|
| 38 |
+
"speculation_length": 0,
|
| 39 |
+
"start_rank_id": 0,
|
| 40 |
+
"target": "trn1",
|
| 41 |
+
"torch_dtype": "float16",
|
| 42 |
+
"tp_degree": 2
|
| 43 |
+
},
|
| 44 |
+
"num_attention_heads": 32,
|
| 45 |
+
"num_experts_per_tok": 2,
|
| 46 |
+
"num_hidden_layers": 2,
|
| 47 |
+
"num_key_value_heads": 8,
|
| 48 |
+
"num_local_experts": 8,
|
| 49 |
+
"output_router_logits": false,
|
| 50 |
+
"rms_norm_eps": 1e-05,
|
| 51 |
+
"rope_theta": 10000.0,
|
| 52 |
+
"router_aux_loss_coef": 0.001,
|
| 53 |
+
"router_jitter_noise": 0.0,
|
| 54 |
+
"sliding_window": 4096,
|
| 55 |
+
"tie_word_embeddings": false,
|
| 56 |
+
"use_cache": true,
|
| 57 |
+
"vocab_size": 32000
|
| 58 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/phi3/yujiepan/phi-4-tiny-random/d2900659f0c3438aa013.json
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "yujiepan/phi-4-tiny-random",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"Phi3ForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"auto_map": {},
|
| 11 |
+
"dtype": "bfloat16",
|
| 12 |
+
"embd_pdrop": 0.0,
|
| 13 |
+
"hidden_act": "silu",
|
| 14 |
+
"hidden_size": 16,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"intermediate_size": 32,
|
| 17 |
+
"max_position_embeddings": 16384,
|
| 18 |
+
"model_type": "phi3",
|
| 19 |
+
"neuron": {
|
| 20 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 21 |
+
"batch_size": 1,
|
| 22 |
+
"capacity_factor": null,
|
| 23 |
+
"checkpoint_id": "yujiepan/phi-4-tiny-random",
|
| 24 |
+
"checkpoint_revision": "18a9a1168dc97ac6d128f811925670c275610f5a",
|
| 25 |
+
"continuous_batching": false,
|
| 26 |
+
"ep_degree": 1,
|
| 27 |
+
"fused_qkv": true,
|
| 28 |
+
"glu_mlp": true,
|
| 29 |
+
"local_ranks_size": 2,
|
| 30 |
+
"max_batch_size": 1,
|
| 31 |
+
"max_context_length": 1024,
|
| 32 |
+
"max_topk": 256,
|
| 33 |
+
"n_active_tokens": 1024,
|
| 34 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 35 |
+
"on_device_sampling": true,
|
| 36 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 37 |
+
"output_logits": false,
|
| 38 |
+
"pp_degree": 1,
|
| 39 |
+
"sequence_length": 1024,
|
| 40 |
+
"speculation_length": 0,
|
| 41 |
+
"start_rank_id": 0,
|
| 42 |
+
"target": "trn1",
|
| 43 |
+
"torch_dtype": "bfloat16",
|
| 44 |
+
"tp_degree": 2
|
| 45 |
+
},
|
| 46 |
+
"num_attention_heads": 2,
|
| 47 |
+
"num_hidden_layers": 2,
|
| 48 |
+
"num_key_value_heads": 1,
|
| 49 |
+
"original_max_position_embeddings": 16384,
|
| 50 |
+
"partial_rotary_factor": 1.0,
|
| 51 |
+
"resid_pdrop": 0.0,
|
| 52 |
+
"rms_norm_eps": 1e-05,
|
| 53 |
+
"rope_scaling": null,
|
| 54 |
+
"rope_theta": 250000,
|
| 55 |
+
"sliding_window": null,
|
| 56 |
+
"tie_word_embeddings": false,
|
| 57 |
+
"use_cache": true,
|
| 58 |
+
"vocab_size": 100352
|
| 59 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/qwen2/yujiepan/qwen2.5-128k-tiny-random/ee9ffbe1ac4a7ad2ee51.json
ADDED
|
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "yujiepan/qwen2.5-128k-tiny-random",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"Qwen2ForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"dtype": "bfloat16",
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 8,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 16,
|
| 14 |
+
"layer_types": [
|
| 15 |
+
"full_attention",
|
| 16 |
+
"full_attention"
|
| 17 |
+
],
|
| 18 |
+
"max_position_embeddings": 32768,
|
| 19 |
+
"max_window_layers": 1,
|
| 20 |
+
"model_type": "qwen2",
|
| 21 |
+
"neuron": {
|
| 22 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 23 |
+
"batch_size": 1,
|
| 24 |
+
"capacity_factor": null,
|
| 25 |
+
"checkpoint_id": "yujiepan/qwen2.5-128k-tiny-random",
|
| 26 |
+
"checkpoint_revision": "c8296d4ca3f87782876d2382fbb6481d1beb8ef0",
|
| 27 |
+
"continuous_batching": false,
|
| 28 |
+
"ep_degree": 1,
|
| 29 |
+
"fused_qkv": false,
|
| 30 |
+
"glu_mlp": true,
|
| 31 |
+
"local_ranks_size": 2,
|
| 32 |
+
"max_batch_size": 1,
|
| 33 |
+
"max_context_length": 1024,
|
| 34 |
+
"max_topk": 256,
|
| 35 |
+
"n_active_tokens": 1024,
|
| 36 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 37 |
+
"on_device_sampling": true,
|
| 38 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 39 |
+
"output_logits": false,
|
| 40 |
+
"pp_degree": 1,
|
| 41 |
+
"sequence_length": 1024,
|
| 42 |
+
"speculation_length": 0,
|
| 43 |
+
"start_rank_id": 0,
|
| 44 |
+
"target": "trn1",
|
| 45 |
+
"torch_dtype": "bfloat16",
|
| 46 |
+
"tp_degree": 2
|
| 47 |
+
},
|
| 48 |
+
"num_attention_heads": 4,
|
| 49 |
+
"num_hidden_layers": 2,
|
| 50 |
+
"num_key_value_heads": 2,
|
| 51 |
+
"rms_norm_eps": 1e-06,
|
| 52 |
+
"rope_scaling": {
|
| 53 |
+
"factor": 4.0,
|
| 54 |
+
"original_max_position_embeddings": 32768,
|
| 55 |
+
"rope_type": "yarn",
|
| 56 |
+
"type": "yarn"
|
| 57 |
+
},
|
| 58 |
+
"rope_theta": 1000000.0,
|
| 59 |
+
"sliding_window": null,
|
| 60 |
+
"tie_word_embeddings": false,
|
| 61 |
+
"use_cache": true,
|
| 62 |
+
"use_sliding_window": false,
|
| 63 |
+
"vocab_size": 152064
|
| 64 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/qwen3_moe/optimum-internal-testing/tiny-random-qwen3_moe/a4bf640c57a2ab4c037e.json
ADDED
|
@@ -0,0 +1,65 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "optimum-internal-testing/tiny-random-qwen3_moe",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"Qwen3MoeForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"decoder_sparse_step": 2,
|
| 11 |
+
"dtype": "float32",
|
| 12 |
+
"head_dim": 32,
|
| 13 |
+
"hidden_act": "silu",
|
| 14 |
+
"hidden_size": 64,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"intermediate_size": 128,
|
| 17 |
+
"max_position_embeddings": 40960,
|
| 18 |
+
"max_window_layers": 1,
|
| 19 |
+
"mlp_only_layers": [],
|
| 20 |
+
"model_type": "qwen3_moe",
|
| 21 |
+
"moe_intermediate_size": 128,
|
| 22 |
+
"neuron": {
|
| 23 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 24 |
+
"batch_size": 1,
|
| 25 |
+
"capacity_factor": null,
|
| 26 |
+
"checkpoint_id": "optimum-internal-testing/tiny-random-qwen3_moe",
|
| 27 |
+
"checkpoint_revision": "e0230be2839556b44b7400a233c73c74b4abb7af",
|
| 28 |
+
"continuous_batching": false,
|
| 29 |
+
"ep_degree": 1,
|
| 30 |
+
"fused_qkv": false,
|
| 31 |
+
"glu_mlp": true,
|
| 32 |
+
"local_ranks_size": 2,
|
| 33 |
+
"max_batch_size": 1,
|
| 34 |
+
"max_context_length": 1024,
|
| 35 |
+
"max_topk": 256,
|
| 36 |
+
"n_active_tokens": 1024,
|
| 37 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 38 |
+
"on_device_sampling": true,
|
| 39 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 40 |
+
"output_logits": false,
|
| 41 |
+
"pp_degree": 1,
|
| 42 |
+
"sequence_length": 1024,
|
| 43 |
+
"speculation_length": 0,
|
| 44 |
+
"start_rank_id": 0,
|
| 45 |
+
"target": "trn1",
|
| 46 |
+
"torch_dtype": "float32",
|
| 47 |
+
"tp_degree": 2
|
| 48 |
+
},
|
| 49 |
+
"norm_topk_prob": true,
|
| 50 |
+
"num_attention_heads": 2,
|
| 51 |
+
"num_experts": 8,
|
| 52 |
+
"num_experts_per_tok": 2,
|
| 53 |
+
"num_hidden_layers": 2,
|
| 54 |
+
"num_key_value_heads": 1,
|
| 55 |
+
"output_router_logits": false,
|
| 56 |
+
"rms_norm_eps": 1e-06,
|
| 57 |
+
"rope_scaling": null,
|
| 58 |
+
"rope_theta": 1000000.0,
|
| 59 |
+
"router_aux_loss_coef": 0.001,
|
| 60 |
+
"sliding_window": null,
|
| 61 |
+
"tie_word_embeddings": true,
|
| 62 |
+
"use_cache": true,
|
| 63 |
+
"use_sliding_window": false,
|
| 64 |
+
"vocab_size": 151936
|
| 65 |
+
}
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/smollm3/HuggingFaceTB/SmolLM3-3B/8adf13bbd08271a796c8.json
ADDED
|
@@ -0,0 +1,134 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "HuggingFaceTB/SmolLM3-3B",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"SmolLM3ForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"dtype": "bfloat16",
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 2048,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 11008,
|
| 15 |
+
"layer_types": [
|
| 16 |
+
"full_attention",
|
| 17 |
+
"full_attention",
|
| 18 |
+
"full_attention",
|
| 19 |
+
"full_attention",
|
| 20 |
+
"full_attention",
|
| 21 |
+
"full_attention",
|
| 22 |
+
"full_attention",
|
| 23 |
+
"full_attention",
|
| 24 |
+
"full_attention",
|
| 25 |
+
"full_attention",
|
| 26 |
+
"full_attention",
|
| 27 |
+
"full_attention",
|
| 28 |
+
"full_attention",
|
| 29 |
+
"full_attention",
|
| 30 |
+
"full_attention",
|
| 31 |
+
"full_attention",
|
| 32 |
+
"full_attention",
|
| 33 |
+
"full_attention",
|
| 34 |
+
"full_attention",
|
| 35 |
+
"full_attention",
|
| 36 |
+
"full_attention",
|
| 37 |
+
"full_attention",
|
| 38 |
+
"full_attention",
|
| 39 |
+
"full_attention",
|
| 40 |
+
"full_attention",
|
| 41 |
+
"full_attention",
|
| 42 |
+
"full_attention",
|
| 43 |
+
"full_attention",
|
| 44 |
+
"full_attention",
|
| 45 |
+
"full_attention",
|
| 46 |
+
"full_attention",
|
| 47 |
+
"full_attention",
|
| 48 |
+
"full_attention",
|
| 49 |
+
"full_attention",
|
| 50 |
+
"full_attention",
|
| 51 |
+
"full_attention"
|
| 52 |
+
],
|
| 53 |
+
"max_position_embeddings": 65536,
|
| 54 |
+
"max_window_layers": 28,
|
| 55 |
+
"mlp_bias": false,
|
| 56 |
+
"model_type": "smollm3",
|
| 57 |
+
"neuron": {
|
| 58 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 59 |
+
"batch_size": 1,
|
| 60 |
+
"capacity_factor": null,
|
| 61 |
+
"checkpoint_id": "HuggingFaceTB/SmolLM3-3B",
|
| 62 |
+
"checkpoint_revision": "a07cc9a04f16550a088caea529712d1d335b0ac1",
|
| 63 |
+
"continuous_batching": false,
|
| 64 |
+
"ep_degree": 1,
|
| 65 |
+
"fused_qkv": true,
|
| 66 |
+
"glu_mlp": true,
|
| 67 |
+
"local_ranks_size": 2,
|
| 68 |
+
"max_batch_size": 1,
|
| 69 |
+
"max_context_length": 1024,
|
| 70 |
+
"max_topk": 256,
|
| 71 |
+
"n_active_tokens": 1024,
|
| 72 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 73 |
+
"on_device_sampling": true,
|
| 74 |
+
"optimum_neuron_version": "0.4.6.dev2",
|
| 75 |
+
"output_logits": false,
|
| 76 |
+
"pp_degree": 1,
|
| 77 |
+
"sequence_length": 1024,
|
| 78 |
+
"speculation_length": 0,
|
| 79 |
+
"start_rank_id": 0,
|
| 80 |
+
"target": "trn1",
|
| 81 |
+
"torch_dtype": "bfloat16",
|
| 82 |
+
"tp_degree": 2
|
| 83 |
+
},
|
| 84 |
+
"no_rope_layer_interval": 4,
|
| 85 |
+
"no_rope_layers": [
|
| 86 |
+
1,
|
| 87 |
+
1,
|
| 88 |
+
1,
|
| 89 |
+
0,
|
| 90 |
+
1,
|
| 91 |
+
1,
|
| 92 |
+
1,
|
| 93 |
+
0,
|
| 94 |
+
1,
|
| 95 |
+
1,
|
| 96 |
+
1,
|
| 97 |
+
0,
|
| 98 |
+
1,
|
| 99 |
+
1,
|
| 100 |
+
1,
|
| 101 |
+
0,
|
| 102 |
+
1,
|
| 103 |
+
1,
|
| 104 |
+
1,
|
| 105 |
+
0,
|
| 106 |
+
1,
|
| 107 |
+
1,
|
| 108 |
+
1,
|
| 109 |
+
0,
|
| 110 |
+
1,
|
| 111 |
+
1,
|
| 112 |
+
1,
|
| 113 |
+
0,
|
| 114 |
+
1,
|
| 115 |
+
1,
|
| 116 |
+
1,
|
| 117 |
+
0,
|
| 118 |
+
1,
|
| 119 |
+
1,
|
| 120 |
+
1,
|
| 121 |
+
0
|
| 122 |
+
],
|
| 123 |
+
"num_attention_heads": 16,
|
| 124 |
+
"num_hidden_layers": 36,
|
| 125 |
+
"num_key_value_heads": 4,
|
| 126 |
+
"pretraining_tp": 2,
|
| 127 |
+
"rms_norm_eps": 1e-06,
|
| 128 |
+
"rope_scaling": null,
|
| 129 |
+
"rope_theta": 5000000.0,
|
| 130 |
+
"sliding_window": null,
|
| 131 |
+
"use_cache": false,
|
| 132 |
+
"use_sliding_window": false,
|
| 133 |
+
"vocab_size": 128256
|
| 134 |
+
}
|
neuronxcc-2.21.33363.0+82129205/MODULE_0d92bb2b1fa1b9d3a3d0+a02c3a36/compile_flags.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
|
neuronxcc-2.21.33363.0+82129205/MODULE_0d92bb2b1fa1b9d3a3d0+a02c3a36/model.done
ADDED
|
File without changes
|
neuronxcc-2.21.33363.0+82129205/MODULE_0d92bb2b1fa1b9d3a3d0+a02c3a36/model.hlo_module.pb
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c0fe8ae6d367e99eb5a27313b6684b26cc8bdc2ffee3d4a8bf5a5251c44beafe
|
| 3 |
+
size 95295
|
neuronxcc-2.21.33363.0+82129205/MODULE_0d92bb2b1fa1b9d3a3d0+a02c3a36/model.neff
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ce71319ab0b4811a8a9085e322f9e9b9529e89b9905a857e3f891fa5fa916041
|
| 3 |
+
size 277504
|
neuronxcc-2.21.33363.0+82129205/MODULE_0d92bb2b1fa1b9d3a3d0+a02c3a36/wrapped_neff.hlo
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a54f900722375339e59690cf409bd8b46face5fa43edf8a67934ad41b283d3b3
|
| 3 |
+
size 288898
|
neuronxcc-2.21.33363.0+82129205/MODULE_142e7dededa7227b3e62+a02c3a36/compile_flags.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
|
neuronxcc-2.21.33363.0+82129205/MODULE_142e7dededa7227b3e62+a02c3a36/model.done
ADDED
|
File without changes
|
neuronxcc-2.21.33363.0+82129205/MODULE_142e7dededa7227b3e62+a02c3a36/model.hlo_module.pb
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0c757db75830e5ca672f8644b5583851422464ea0c30bbb38f612bb55427b554
|
| 3 |
+
size 680451
|
neuronxcc-2.21.33363.0+82129205/MODULE_142e7dededa7227b3e62+a02c3a36/model.neff
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14fd4ec0d1dd57153a23287071f616db83c0660e597a2aa7ee9b9ca9436bfdad
|
| 3 |
+
size 543744
|
neuronxcc-2.21.33363.0+82129205/MODULE_142e7dededa7227b3e62+a02c3a36/wrapped_neff.hlo
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ee9c92eab70cf570b30720938093022dfd3ceca77479abd878b4078318d07cc
|
| 3 |
+
size 563245
|
neuronxcc-2.21.33363.0+82129205/MODULE_15c98fdcb7e9199cea41+a02c3a36/compile_flags.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
|
neuronxcc-2.21.33363.0+82129205/MODULE_15c98fdcb7e9199cea41+a02c3a36/model.done
ADDED
|
File without changes
|
neuronxcc-2.21.33363.0+82129205/MODULE_15c98fdcb7e9199cea41+a02c3a36/model.hlo_module.pb
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3dbba0a1af2a53f1fb42fe1ace2a91f1ded9c5df6592cd34424abaaf8b82bde9
|
| 3 |
+
size 84950
|
neuronxcc-2.21.33363.0+82129205/MODULE_15c98fdcb7e9199cea41+a02c3a36/model.neff
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:135c3b437ec9ed62dda890966116857c44f95358cc44e5a7c436a158ff5d20d0
|
| 3 |
+
size 246784
|
neuronxcc-2.21.33363.0+82129205/MODULE_15c98fdcb7e9199cea41+a02c3a36/wrapped_neff.hlo
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:97b50ecf8ea86471998cc1c80b24874535b449cdb9408578b65b5c4fc1298089
|
| 3 |
+
size 254967
|
neuronxcc-2.21.33363.0+82129205/MODULE_2dd5e80c893e94b8c53c+677eeb9d/compile_flags.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/speculation_model/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
|
neuronxcc-2.21.33363.0+82129205/MODULE_2dd5e80c893e94b8c53c+677eeb9d/model.done
ADDED
|
File without changes
|
neuronxcc-2.21.33363.0+82129205/MODULE_2dd5e80c893e94b8c53c+677eeb9d/model.hlo_module.pb
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1aed48d406d79d5d3e739ae6fa23059cd3a200ff97ed068e452cf2aded1bd09
|
| 3 |
+
size 453530
|
neuronxcc-2.21.33363.0+82129205/MODULE_2dd5e80c893e94b8c53c+677eeb9d/model.neff
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec74b05853ca38742f004b3327e17e41e436de1c7f28d415e914459d99ed1491
|
| 3 |
+
size 3984384
|
neuronxcc-2.21.33363.0+82129205/MODULE_36e9f4ac79a30a03d15c+a02c3a36/compile_flags.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
|
neuronxcc-2.21.33363.0+82129205/MODULE_36e9f4ac79a30a03d15c+a02c3a36/model.done
ADDED
|
File without changes
|
neuronxcc-2.21.33363.0+82129205/MODULE_36e9f4ac79a30a03d15c+a02c3a36/model.hlo_module.pb
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5909b61d8f3268e4bc41b2411af6e4d216f5f50670988c4e9e320b4b57e74e30
|
| 3 |
+
size 403385
|
neuronxcc-2.21.33363.0+82129205/MODULE_36e9f4ac79a30a03d15c+a02c3a36/model.neff
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:93bf36e95d20a0e77475315ef8d1cdc93d47cf854e650e302ec18894fdadc619
|
| 3 |
+
size 2018304
|
neuronxcc-2.21.33363.0+82129205/MODULE_36e9f4ac79a30a03d15c+a02c3a36/wrapped_neff.hlo
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:79d95ad2132ad479e7fda69e0d1c63a92b9d981d9479c71dc2861c5556691bbe
|
| 3 |
+
size 2111217
|
neuronxcc-2.21.33363.0+82129205/MODULE_3ee7521f420173f2d5db+80d05c3f/compile_flags.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
["--target=trn1", "--enable-saturate-infinity", "--enable-mixed-precision-accumulation", "--model-type", "transformer", "-O1", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2", "--auto-cast=none", "--internal-enable-dge-levels", "vector_dynamic_offsets", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
|
neuronxcc-2.21.33363.0+82129205/MODULE_3ee7521f420173f2d5db+80d05c3f/model.done
ADDED
|
File without changes
|
neuronxcc-2.21.33363.0+82129205/MODULE_3ee7521f420173f2d5db+80d05c3f/model.hlo_module.pb
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d1b86d2eda6afe0bf03aa1f755e61cf0930adbd115277b035a5ba71919cf994
|
| 3 |
+
size 74422
|
neuronxcc-2.21.33363.0+82129205/MODULE_3ee7521f420173f2d5db+80d05c3f/model.neff
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d927a05395fc0145315368108eeb2830e201f19c925609c1d473a54278458440
|
| 3 |
+
size 277504
|
neuronxcc-2.21.33363.0+82129205/MODULE_3ee7521f420173f2d5db+80d05c3f/wrapped_neff.hlo
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:af695cfdf89a2a5f91a8fe5bec776c2bc952e3f98a15658b400d17d7b129904b
|
| 3 |
+
size 289434
|