Synchronizing local compiler cache.
Browse files- .gitattributes +4 -0
- neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.3.0.dev2/llama/unsloth/Llama-3.2-1B-Instruct/0a8784a00d0c8111b947.json +78 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_1fc614255e6a9d7b109d+841d78e1/compile_flags.json +1 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_1fc614255e6a9d7b109d+841d78e1/model.done +0 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_1fc614255e6a9d7b109d+841d78e1/model.hlo_module.pb +3 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_1fc614255e6a9d7b109d+841d78e1/model.neff +3 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_bebc77fa7efe716566fe+165e9558/compile_flags.json +1 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_bebc77fa7efe716566fe+165e9558/model.done +0 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_bebc77fa7efe716566fe+165e9558/model.hlo_module.pb +3 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_bebc77fa7efe716566fe+165e9558/model.neff +3 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_bebc77fa7efe716566fe+165e9558/wrapped_neff.hlo +3 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_fc12e674baf93fa7e151+bfe5714b/compile_flags.json +1 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_fc12e674baf93fa7e151+bfe5714b/model.done +0 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_fc12e674baf93fa7e151+bfe5714b/model.hlo_module.pb +3 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_fc12e674baf93fa7e151+bfe5714b/model.neff +3 -0
.gitattributes
CHANGED
@@ -3109,3 +3109,7 @@ neuronxcc-2.17.194.0+d312836f/MODULE_7a301946be6cceeea745+bfe5714b/model.neff fi
|
|
3109 |
neuronxcc-2.17.194.0+d312836f/MODULE_7ef58b1e1f63c382674a+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
|
3110 |
neuronxcc-2.17.194.0+d312836f/MODULE_efd8799e097b083a4ed2+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
|
3111 |
neuronxcc-2.17.194.0+d312836f/MODULE_efd8799e097b083a4ed2+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
3109 |
neuronxcc-2.17.194.0+d312836f/MODULE_7ef58b1e1f63c382674a+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
|
3110 |
neuronxcc-2.17.194.0+d312836f/MODULE_efd8799e097b083a4ed2+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
|
3111 |
neuronxcc-2.17.194.0+d312836f/MODULE_efd8799e097b083a4ed2+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
3112 |
+
neuronxcc-2.17.194.0+d312836f/MODULE_1fc614255e6a9d7b109d+841d78e1/model.neff filter=lfs diff=lfs merge=lfs -text
|
3113 |
+
neuronxcc-2.17.194.0+d312836f/MODULE_bebc77fa7efe716566fe+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
|
3114 |
+
neuronxcc-2.17.194.0+d312836f/MODULE_bebc77fa7efe716566fe+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
3115 |
+
neuronxcc-2.17.194.0+d312836f/MODULE_fc12e674baf93fa7e151+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
|
neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.3.0.dev2/llama/unsloth/Llama-3.2-1B-Instruct/0a8784a00d0c8111b947.json
ADDED
@@ -0,0 +1,78 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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 |
+
"head_dim": 64,
|
11 |
+
"hidden_act": "silu",
|
12 |
+
"hidden_size": 2048,
|
13 |
+
"initializer_range": 0.02,
|
14 |
+
"intermediate_size": 8192,
|
15 |
+
"max_position_embeddings": 131072,
|
16 |
+
"mlp_bias": false,
|
17 |
+
"model_type": "llama",
|
18 |
+
"neuron": {
|
19 |
+
"_serialized_key": "NxDNeuronConfig",
|
20 |
+
"async_mode": false,
|
21 |
+
"attn_kernel_enabled": false,
|
22 |
+
"batch_size": 1,
|
23 |
+
"capacity_factor": null,
|
24 |
+
"cc_pipeline_tiling_factor": 2,
|
25 |
+
"checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
|
26 |
+
"checkpoint_revision": null,
|
27 |
+
"continuous_batching": false,
|
28 |
+
"enable_bucketing": false,
|
29 |
+
"ep_degree": 1,
|
30 |
+
"flash_decoding_enabled": false,
|
31 |
+
"fused_qkv": false,
|
32 |
+
"glu_mlp": true,
|
33 |
+
"is_chunked_prefill": false,
|
34 |
+
"local_ranks_size": 2,
|
35 |
+
"logical_nc_config": 1,
|
36 |
+
"max_batch_size": 1,
|
37 |
+
"max_context_length": 4096,
|
38 |
+
"max_topk": 256,
|
39 |
+
"mlp_kernel_enabled": false,
|
40 |
+
"mlp_kernel_fuse_residual_add": false,
|
41 |
+
"n_active_tokens": 4096,
|
42 |
+
"neuronxcc_version": "2.17.194.0+d312836f",
|
43 |
+
"num_cores_per_group": 1,
|
44 |
+
"on_device_sampling": false,
|
45 |
+
"optimum_neuron_version": "0.3.0.dev2",
|
46 |
+
"output_logits": false,
|
47 |
+
"padding_side": "right",
|
48 |
+
"pp_degree": 1,
|
49 |
+
"qk_layernorm": false,
|
50 |
+
"qkv_kernel_enabled": false,
|
51 |
+
"rpl_reduce_dtype": "bfloat16",
|
52 |
+
"sequence_length": 4096,
|
53 |
+
"sequence_parallel_enabled": false,
|
54 |
+
"speculation_length": 5,
|
55 |
+
"start_rank_id": 0,
|
56 |
+
"target": null,
|
57 |
+
"torch_dtype": "bfloat16",
|
58 |
+
"tp_degree": 2,
|
59 |
+
"vocab_parallel": false
|
60 |
+
},
|
61 |
+
"num_attention_heads": 32,
|
62 |
+
"num_hidden_layers": 16,
|
63 |
+
"num_key_value_heads": 8,
|
64 |
+
"pretraining_tp": 1,
|
65 |
+
"rms_norm_eps": 1e-05,
|
66 |
+
"rope_scaling": {
|
67 |
+
"factor": 32.0,
|
68 |
+
"high_freq_factor": 4.0,
|
69 |
+
"low_freq_factor": 1.0,
|
70 |
+
"original_max_position_embeddings": 8192,
|
71 |
+
"rope_type": "llama3"
|
72 |
+
},
|
73 |
+
"rope_theta": 500000.0,
|
74 |
+
"tie_word_embeddings": true,
|
75 |
+
"unsloth_fixed": true,
|
76 |
+
"use_cache": true,
|
77 |
+
"vocab_size": 128256
|
78 |
+
}
|
neuronxcc-2.17.194.0+d312836f/MODULE_1fc614255e6a9d7b109d+841d78e1/compile_flags.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
"--auto-cast=none --model-type=transformer --tensorizer-options='--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-dge-dma --vectorize-strided-dma ' -O2 --internal-num-neuroncores-per-sengine=1 --logfile=/tmp/nxd_model/speculation_model/_tp0_bk0/log-neuron-cc.txt --enable-internal-neff-wrapper"
|
neuronxcc-2.17.194.0+d312836f/MODULE_1fc614255e6a9d7b109d+841d78e1/model.done
ADDED
File without changes
|
neuronxcc-2.17.194.0+d312836f/MODULE_1fc614255e6a9d7b109d+841d78e1/model.hlo_module.pb
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:33fdc7be286618543839c1656a87744f5365657343413672bd10a29d7530cac8
|
3 |
+
size 1032672
|
neuronxcc-2.17.194.0+d312836f/MODULE_1fc614255e6a9d7b109d+841d78e1/model.neff
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:843df4540a967409c8538056292b7e9a1245f6f8929c2ac0138d40e35fcea505
|
3 |
+
size 2704384
|
neuronxcc-2.17.194.0+d312836f/MODULE_bebc77fa7efe716566fe+165e9558/compile_flags.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
"--auto-cast=none --model-type=transformer --tensorizer-options='--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-dge-dma --vectorize-strided-dma ' -O2 --internal-num-neuroncores-per-sengine=1 --logfile=/tmp/nxd_model/token_generation_model/_tp0_bk0/log-neuron-cc.txt --enable-internal-neff-wrapper"
|
neuronxcc-2.17.194.0+d312836f/MODULE_bebc77fa7efe716566fe+165e9558/model.done
ADDED
File without changes
|
neuronxcc-2.17.194.0+d312836f/MODULE_bebc77fa7efe716566fe+165e9558/model.hlo_module.pb
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4ffadf090f1e5b25d004205a163f73aef66771ba6ea46cc1d4c8fe62886c7f38
|
3 |
+
size 954337
|
neuronxcc-2.17.194.0+d312836f/MODULE_bebc77fa7efe716566fe+165e9558/model.neff
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9c69cf3b769a6aaa7e1dd37d6f0b89b0d3ad47f495c53daf61bbde90943e7dd1
|
3 |
+
size 1526784
|
neuronxcc-2.17.194.0+d312836f/MODULE_bebc77fa7efe716566fe+165e9558/wrapped_neff.hlo
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7c5e2290eb347e96386b6e03f93c43d03bbb81fc3ece691ceb9816dc73e88956
|
3 |
+
size 1613754
|
neuronxcc-2.17.194.0+d312836f/MODULE_fc12e674baf93fa7e151+bfe5714b/compile_flags.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
"--auto-cast=none --model-type=transformer --tensorizer-options='--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-dge-dma --vectorize-strided-dma ' -O2 --internal-num-neuroncores-per-sengine=1 --logfile=/tmp/nxd_model/context_encoding_model/_tp0_bk0/log-neuron-cc.txt"
|
neuronxcc-2.17.194.0+d312836f/MODULE_fc12e674baf93fa7e151+bfe5714b/model.done
ADDED
File without changes
|
neuronxcc-2.17.194.0+d312836f/MODULE_fc12e674baf93fa7e151+bfe5714b/model.hlo_module.pb
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9abcce14f0414e1c45e61f4757dd660466ac74ca33bd87b99b4eb958bdacc981
|
3 |
+
size 737474
|
neuronxcc-2.17.194.0+d312836f/MODULE_fc12e674baf93fa7e151+bfe5714b/model.neff
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1586289fd962cdcec0d689bb99a71f6dd2c6c7afd8535e64f5eff7ac20443d47
|
3 |
+
size 31724544
|