Commit
6bdf9aa
·
verified ·
1 Parent(s): d02a489

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -3556,3 +3556,7 @@ neuronxcc-2.17.194.0+d312836f/MODULE_1e33dc3a7f5459878e36+bfe5714b/model.neff fi
3556
  neuronxcc-2.17.194.0+d312836f/MODULE_531c872df3c2c462cafe+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
3557
  neuronxcc-2.17.194.0+d312836f/MODULE_531c872df3c2c462cafe+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3558
  neuronxcc-2.17.194.0+d312836f/MODULE_9430ff6a35fc239154d1+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
3556
  neuronxcc-2.17.194.0+d312836f/MODULE_531c872df3c2c462cafe+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
3557
  neuronxcc-2.17.194.0+d312836f/MODULE_531c872df3c2c462cafe+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3558
  neuronxcc-2.17.194.0+d312836f/MODULE_9430ff6a35fc239154d1+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
3559
+ neuronxcc-2.17.194.0+d312836f/MODULE_69efb951886262c3c792+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
3560
+ neuronxcc-2.17.194.0+d312836f/MODULE_69efb951886262c3c792+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3561
+ neuronxcc-2.17.194.0+d312836f/MODULE_a642e4644f4dd00e7e7c+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
3562
+ neuronxcc-2.17.194.0+d312836f/MODULE_bde59c9c06380a392ee9+841d78e1/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.3.0.dev4/llama/unsloth/Llama-3.2-1B-Instruct/fcb54286a292bb55ad35.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.dev4",
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_69efb951886262c3c792+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_69efb951886262c3c792+165e9558/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_69efb951886262c3c792+165e9558/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:72a1e539c28e382d34b7a5475aa645008dc8a8e023164b353d28db8d1b4da571
3
+ size 379979
neuronxcc-2.17.194.0+d312836f/MODULE_69efb951886262c3c792+165e9558/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:090761ff224eac1de4e7b0159acdb3b00bc741d7b37119f2f057bf1022d9ec4d
3
+ size 1516544
neuronxcc-2.17.194.0+d312836f/MODULE_69efb951886262c3c792+165e9558/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca6f0dee7e15b4d696edaf274c92c140cc4181deec059325ebfab01f34007407
3
+ size 1603514
neuronxcc-2.17.194.0+d312836f/MODULE_a642e4644f4dd00e7e7c+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_a642e4644f4dd00e7e7c+bfe5714b/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_a642e4644f4dd00e7e7c+bfe5714b/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fa67d9d1db2ddd4e707b1e33e9225df31ee9bb6f6dd8075dc771cbf5c755332
3
+ size 375758
neuronxcc-2.17.194.0+d312836f/MODULE_a642e4644f4dd00e7e7c+bfe5714b/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a0f831148896502150bf4230578e178b6bb933b09599bd37bf94ffbc4ce75de
3
+ size 31693824
neuronxcc-2.17.194.0+d312836f/MODULE_bde59c9c06380a392ee9+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_bde59c9c06380a392ee9+841d78e1/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_bde59c9c06380a392ee9+841d78e1/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04c4672203ab6f881de528a93df2036354397fee34821eead69c319ac2ce024b
3
+ size 415162
neuronxcc-2.17.194.0+d312836f/MODULE_bde59c9c06380a392ee9+841d78e1/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac1117116e28b65af239c3fbbd4eda3f64b4bde5d5ec81b58ce7afb401fdfe15
3
+ size 2694144