Hugging Face's logo

INCModel3
/
FLUX.2-dev-MXFP4-RTN-AutoRound

Text-to-Image
Diffusers
Safetensors
Flux2Pipeline
quantized
mxfp4
autoround
diffusion
autoquant-agent
Model card Files
xet
 Community
lvkaokao commited on
Commit
59fb46e
·
verified ·
1 Parent(s): 5606291

Upload quantized model FLUX.2-dev-MXFP4-RTN-AutoRound

Browse files
Files changed (20) hide show
  1. .gitattributes +1 -0
  2. model_index.json +25 -0
  3. scheduler/scheduler_config.json +18 -0
  4. text_encoder/config.json +59 -0
  5. text_encoder/generation_config.json +8 -0
  6. text_encoder/model.safetensors +3 -0
  7. tokenizer/chat_template.jinja +51 -0
  8. tokenizer/processor_config.json +33 -0
  9. tokenizer/tokenizer.json +3 -0
  10. tokenizer/tokenizer_config.json +16 -0
  11. transformer/config.json +45 -0
  12. transformer/model-00001-of-00005.safetensors +3 -0
  13. transformer/model-00002-of-00005.safetensors +3 -0
  14. transformer/model-00003-of-00005.safetensors +3 -0
  15. transformer/model-00004-of-00005.safetensors +3 -0
  16. transformer/model-00005-of-00005.safetensors +3 -0
  17. transformer/model.safetensors.index.json +533 -0
  18. transformer/quantization_config.json +18 -0
  19. vae/config.json +41 -0
  20. vae/diffusion_pytorch_model.safetensors +3 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
model_index.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "Flux2Pipeline",
3
+ "_diffusers_version": "0.40.0",
4
+ "_name_or_path": "/models/FLUX.2-dev",
5
+ "scheduler": [
6
+ "diffusers",
7
+ "FlowMatchEulerDiscreteScheduler"
8
+ ],
9
+ "text_encoder": [
10
+ "transformers",
11
+ "Mistral3ForConditionalGeneration"
12
+ ],
13
+ "tokenizer": [
14
+ "transformers",
15
+ "PixtralProcessor"
16
+ ],
17
+ "transformer": [
18
+ "diffusers",
19
+ "Flux2Transformer2DModel"
20
+ ],
21
+ "vae": [
22
+ "diffusers",
23
+ "AutoencoderKLFlux2"
24
+ ]
25
+ }
scheduler/scheduler_config.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "FlowMatchEulerDiscreteScheduler",
3
+ "_diffusers_version": "0.40.0",
4
+ "base_image_seq_len": 256,
5
+ "base_shift": 0.5,
6
+ "invert_sigmas": false,
7
+ "max_image_seq_len": 4096,
8
+ "max_shift": 1.15,
9
+ "num_train_timesteps": 1000,
10
+ "shift": 3.0,
11
+ "shift_terminal": null,
12
+ "stochastic_sampling": false,
13
+ "time_shift_type": "exponential",
14
+ "use_beta_sigmas": false,
15
+ "use_dynamic_shifting": true,
16
+ "use_exponential_sigmas": false,
17
+ "use_karras_sigmas": false
18
+ }
text_encoder/config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Mistral3ForConditionalGeneration"
4
+ ],
5
+ "dtype": "bfloat16",
6
+ "image_token_index": 10,
7
+ "model_type": "mistral3",
8
+ "multimodal_projector_bias": false,
9
+ "projector_hidden_act": "gelu",
10
+ "spatial_merge_size": 2,
11
+ "text_config": {
12
+ "attention_dropout": 0.0,
13
+ "bos_token_id": 1,
14
+ "dtype": "bfloat16",
15
+ "eos_token_id": 2,
16
+ "head_dim": 128,
17
+ "hidden_act": "silu",
18
+ "hidden_size": 5120,
19
+ "initializer_range": 0.02,
20
+ "intermediate_size": 32768,
21
+ "max_position_embeddings": 131072,
22
+ "model_type": "mistral",
23
+ "num_attention_heads": 32,
24
+ "num_hidden_layers": 40,
25
+ "num_key_value_heads": 8,
26
+ "pad_token_id": null,
27
+ "rms_norm_eps": 1e-05,
28
+ "rope_parameters": {
29
+ "rope_theta": 1000000000.0,
30
+ "rope_type": "default"
31
+ },
32
+ "sliding_window": null,
33
+ "tie_word_embeddings": false,
34
+ "use_cache": true,
35
+ "vocab_size": 131072
36
+ },
37
+ "tie_word_embeddings": true,
38
+ "transformers_version": "5.14.1",
39
+ "vision_config": {
40
+ "attention_dropout": 0.0,
41
+ "dtype": "bfloat16",
42
+ "head_dim": 64,
43
+ "hidden_act": "silu",
44
+ "hidden_size": 1024,
45
+ "image_size": 1540,
46
+ "initializer_range": 0.02,
47
+ "intermediate_size": 4096,
48
+ "model_type": "pixtral",
49
+ "num_attention_heads": 16,
50
+ "num_channels": 3,
51
+ "num_hidden_layers": 24,
52
+ "patch_size": 14,
53
+ "rope_parameters": {
54
+ "rope_theta": 10000.0,
55
+ "rope_type": "default"
56
+ }
57
+ },
58
+ "vision_feature_layer": -1
59
+ }
text_encoder/generation_config.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "do_sample": true,
5
+ "eos_token_id": 2,
6
+ "temperature": 0.15,
7
+ "transformers_version": "5.14.1"
8
+ }
text_encoder/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:06405bef994b95760c931bdacc37303ef5c4f116890e436e275f591241e10c12
3
+ size 48022801424
tokenizer/chat_template.jinja ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- set today = strftime_now("%Y-%m-%d") %}
2
+ {%- set default_system_message = "You are Mistral Small 3, a Large Language Model (LLM) created by Mistral AI, a French startup headquartered in Paris.\nYour knowledge base was last updated on 2023-10-01. The current date is " + today + ".\n\nWhen you're not sure about some information, you say that you don't have the information and don't make up anything.\nIf the user's question is not clear, ambiguous, or does not provide enough context for you to accurately answer the question, you do not try to answer it right away and you rather ask the user to clarify their request (e.g. \"What are some good restaurants around me?\" => \"Where are you?\" or \"When is the next flight to Tokyo\" => \"Where do you travel from?\")" %}
3
+
4
+ {{- bos_token }}
5
+
6
+ {%- if messages[0]['role'] == 'system' %}
7
+ {%- if messages[0]['content'] is string %}
8
+ {%- set system_message = messages[0]['content'] %}
9
+ {%- else %}
10
+ {%- set system_message = messages[0]['content'][0]['text'] %}
11
+ {%- endif %}
12
+ {%- set loop_messages = messages[1:] %}
13
+ {%- else %}
14
+ {%- set system_message = default_system_message %}
15
+ {%- set loop_messages = messages %}
16
+ {%- endif %}
17
+ {{- '[SYSTEM_PROMPT]' + system_message + '[/SYSTEM_PROMPT]' }}
18
+
19
+ {%- for message in loop_messages %}
20
+ {%- if message['role'] == 'user' %}
21
+ {%- if message['content'] is string %}
22
+ {{- '[INST]' + message['content'] + '[/INST]' }}
23
+ {%- else %}
24
+ {{- '[INST]' }}
25
+ {%- for block in message['content'] %}
26
+ {%- if block['type'] == 'text' %}
27
+ {{- block['text'] }}
28
+ {%- elif block['type'] in ['image', 'image_url'] %}
29
+ {{- '[IMG]' }}
30
+ {%- else %}
31
+ {{- raise_exception('Only text and image blocks are supported in message content!') }}
32
+ {%- endif %}
33
+ {%- endfor %}
34
+ {{- '[/INST]' }}
35
+ {%- endif %}
36
+ {%- elif message['role'] == 'system' %}
37
+ {%- if message['content'] is string %}
38
+ {{- '[SYSTEM_PROMPT]' + message['content'] + '[/SYSTEM_PROMPT]' }}
39
+ {%- else %}
40
+ {{- '[SYSTEM_PROMPT]' + message['content'][0]['text'] + '[/SYSTEM_PROMPT]' }}
41
+ {%- endif %}
42
+ {%- elif message['role'] == 'assistant' %}
43
+ {%- if message['content'] is string %}
44
+ {{- message['content'] + eos_token }}
45
+ {%- else %}
46
+ {{- message['content'][0]['text'] + eos_token }}
47
+ {%- endif %}
48
+ {%- else %}
49
+ {{- raise_exception('Only user, system and assistant roles are supported!') }}
50
+ {%- endif %}
51
+ {%- endfor %}
tokenizer/processor_config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_break_token": "[IMG_BREAK]",
3
+ "image_end_token": "[IMG_END]",
4
+ "image_processor": {
5
+ "data_format": "channels_first",
6
+ "default_to_square": true,
7
+ "do_convert_rgb": true,
8
+ "do_normalize": true,
9
+ "do_rescale": true,
10
+ "do_resize": true,
11
+ "image_mean": [
12
+ 0.48145466,
13
+ 0.4578275,
14
+ 0.40821073
15
+ ],
16
+ "image_processor_type": "PixtralImageProcessor",
17
+ "image_std": [
18
+ 0.26862954,
19
+ 0.26130258,
20
+ 0.27577711
21
+ ],
22
+ "patch_size": 14,
23
+ "resample": 3,
24
+ "rescale_factor": 0.00392156862745098,
25
+ "size": {
26
+ "longest_edge": 1540
27
+ }
28
+ },
29
+ "image_token": "[IMG]",
30
+ "patch_size": 14,
31
+ "processor_class": "PixtralProcessor",
32
+ "spatial_merge_size": 2
33
+ }
tokenizer/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b76085f9923309d873994d444989f7eb6ec074b06f25b58f1e8d7b7741070949
3
+ size 17078037
tokenizer/tokenizer_config.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": null,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<s>",
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "</s>",
7
+ "is_local": true,
8
+ "legacy": true,
9
+ "local_files_only": false,
10
+ "model_max_length": 1000000000000000019884624838656,
11
+ "pad_token": "<pad>",
12
+ "processor_class": "PixtralProcessor",
13
+ "tokenizer_class": "TokenizersBackend",
14
+ "unk_token": "<unk>",
15
+ "use_default_system_prompt": false
16
+ }
transformer/config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "Flux2Transformer2DModel",
3
+ "_diffusers_version": "0.36.0.dev0",
4
+ "_name_or_path": "/models/FLUX.2-dev/transformer",
5
+ "_use_default_values": [
6
+ "guidance_embeds"
7
+ ],
8
+ "attention_head_dim": 128,
9
+ "axes_dims_rope": [
10
+ 32,
11
+ 32,
12
+ 32,
13
+ 32
14
+ ],
15
+ "eps": 1e-06,
16
+ "guidance_embeds": true,
17
+ "in_channels": 128,
18
+ "joint_attention_dim": 15360,
19
+ "mlp_ratio": 3.0,
20
+ "num_attention_heads": 48,
21
+ "num_layers": 8,
22
+ "num_single_layers": 48,
23
+ "out_channels": null,
24
+ "patch_size": 1,
25
+ "quantization_config": {
26
+ "act_bits": 4,
27
+ "act_data_type": "mx_fp",
28
+ "act_dynamic": true,
29
+ "act_group_size": 32,
30
+ "act_sym": true,
31
+ "autoround_version": "0.15.0",
32
+ "bits": 4,
33
+ "block_name_to_quantize": "transformer_blocks,single_transformer_blocks",
34
+ "data_type": "mx_fp",
35
+ "enable_quanted_input": false,
36
+ "group_size": 32,
37
+ "packing_format": "auto_round:llm_compressor",
38
+ "quant_method": "auto-round",
39
+ "static_attention_granularity": "tensor",
40
+ "static_kv_granularity": "tensor",
41
+ "sym": true
42
+ },
43
+ "rope_theta": 2000,
44
+ "timestep_guidance_channels": 256
45
+ }
transformer/model-00001-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d399be26150e24d73d27f6c2231eb6395960ef0a7c85ad36b4674f6247dd2e4a
3
+ size 4171271432
transformer/model-00002-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:41e49175ecb0eaa964db4590945b7845ba9a2c6ccd6894ca64c63718b95937f8
3
+ size 4171255536
transformer/model-00003-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:acaec8d3007cfee1743306ee2a5baf88dd518ed007a699fb68f41617aeeba7d0
3
+ size 4171255600
transformer/model-00004-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16ff35add63be8331416c3745b234d8ca9588322178aff502079ecf8d4ddaf7c
3
+ size 4253044176
transformer/model-00005-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:df8af502e8d85edf786fec59fce42d88700f9ff1873b428081f2036b9262eac8
3
+ size 1550844792
transformer/model.safetensors.index.json ADDED
@@ -0,0 +1,533 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "format": "safetensors",
4
+ "total_shards": 5,
5
+ "total_parameters": 17501274112,
6
+ "total_size": 18317606912
7
+ },
8
+ "weight_map": {
9
+ "transformer_blocks.0.attn.norm_q.weight": "model-00001-of-00005.safetensors",
10
+ "transformer_blocks.0.attn.norm_k.weight": "model-00001-of-00005.safetensors",
11
+ "transformer_blocks.0.attn.norm_added_q.weight": "model-00001-of-00005.safetensors",
12
+ "transformer_blocks.0.attn.norm_added_k.weight": "model-00001-of-00005.safetensors",
13
+ "transformer_blocks.0.attn.to_q.weight_packed": "model-00001-of-00005.safetensors",
14
+ "transformer_blocks.0.attn.to_q.weight_scale": "model-00001-of-00005.safetensors",
15
+ "transformer_blocks.0.attn.to_k.weight_packed": "model-00001-of-00005.safetensors",
16
+ "transformer_blocks.0.attn.to_k.weight_scale": "model-00001-of-00005.safetensors",
17
+ "transformer_blocks.0.attn.to_v.weight_packed": "model-00001-of-00005.safetensors",
18
+ "transformer_blocks.0.attn.to_v.weight_scale": "model-00001-of-00005.safetensors",
19
+ "transformer_blocks.0.attn.to_out.0.weight_packed": "model-00001-of-00005.safetensors",
20
+ "transformer_blocks.0.attn.to_out.0.weight_scale": "model-00001-of-00005.safetensors",
21
+ "transformer_blocks.0.attn.add_q_proj.weight_packed": "model-00001-of-00005.safetensors",
22
+ "transformer_blocks.0.attn.add_q_proj.weight_scale": "model-00001-of-00005.safetensors",
23
+ "transformer_blocks.0.attn.add_k_proj.weight_packed": "model-00001-of-00005.safetensors",
24
+ "transformer_blocks.0.attn.add_k_proj.weight_scale": "model-00001-of-00005.safetensors",
25
+ "transformer_blocks.0.attn.add_v_proj.weight_packed": "model-00001-of-00005.safetensors",
26
+ "transformer_blocks.0.attn.add_v_proj.weight_scale": "model-00001-of-00005.safetensors",
27
+ "transformer_blocks.0.attn.to_add_out.weight_packed": "model-00001-of-00005.safetensors",
28
+ "transformer_blocks.0.attn.to_add_out.weight_scale": "model-00001-of-00005.safetensors",
29
+ "transformer_blocks.0.ff.linear_in.weight_packed": "model-00001-of-00005.safetensors",
30
+ "transformer_blocks.0.ff.linear_in.weight_scale": "model-00001-of-00005.safetensors",
31
+ "transformer_blocks.0.ff.linear_out.weight_packed": "model-00001-of-00005.safetensors",
32
+ "transformer_blocks.0.ff.linear_out.weight_scale": "model-00001-of-00005.safetensors",
33
+ "transformer_blocks.0.ff_context.linear_in.weight_packed": "model-00001-of-00005.safetensors",
34
+ "transformer_blocks.0.ff_context.linear_in.weight_scale": "model-00001-of-00005.safetensors",
35
+ "transformer_blocks.0.ff_context.linear_out.weight_packed": "model-00001-of-00005.safetensors",
36
+ "transformer_blocks.0.ff_context.linear_out.weight_scale": "model-00001-of-00005.safetensors",
37
+ "transformer_blocks.1.attn.norm_q.weight": "model-00001-of-00005.safetensors",
38
+ "transformer_blocks.1.attn.norm_k.weight": "model-00001-of-00005.safetensors",
39
+ "transformer_blocks.1.attn.norm_added_q.weight": "model-00001-of-00005.safetensors",
40
+ "transformer_blocks.1.attn.norm_added_k.weight": "model-00001-of-00005.safetensors",
41
+ "transformer_blocks.1.attn.to_q.weight_packed": "model-00001-of-00005.safetensors",
42
+ "transformer_blocks.1.attn.to_q.weight_scale": "model-00001-of-00005.safetensors",
43
+ "transformer_blocks.1.attn.to_k.weight_packed": "model-00001-of-00005.safetensors",
44
+ "transformer_blocks.1.attn.to_k.weight_scale": "model-00001-of-00005.safetensors",
45
+ "transformer_blocks.1.attn.to_v.weight_packed": "model-00001-of-00005.safetensors",
46
+ "transformer_blocks.1.attn.to_v.weight_scale": "model-00001-of-00005.safetensors",
47
+ "transformer_blocks.1.attn.to_out.0.weight_packed": "model-00001-of-00005.safetensors",
48
+ "transformer_blocks.1.attn.to_out.0.weight_scale": "model-00001-of-00005.safetensors",
49
+ "transformer_blocks.1.attn.add_q_proj.weight_packed": "model-00001-of-00005.safetensors",
50
+ "transformer_blocks.1.attn.add_q_proj.weight_scale": "model-00001-of-00005.safetensors",
51
+ "transformer_blocks.1.attn.add_k_proj.weight_packed": "model-00001-of-00005.safetensors",
52
+ "transformer_blocks.1.attn.add_k_proj.weight_scale": "model-00001-of-00005.safetensors",
53
+ "transformer_blocks.1.attn.add_v_proj.weight_packed": "model-00001-of-00005.safetensors",
54
+ "transformer_blocks.1.attn.add_v_proj.weight_scale": "model-00001-of-00005.safetensors",
55
+ "transformer_blocks.1.attn.to_add_out.weight_packed": "model-00001-of-00005.safetensors",
56
+ "transformer_blocks.1.attn.to_add_out.weight_scale": "model-00001-of-00005.safetensors",
57
+ "transformer_blocks.1.ff.linear_in.weight_packed": "model-00001-of-00005.safetensors",
58
+ "transformer_blocks.1.ff.linear_in.weight_scale": "model-00001-of-00005.safetensors",
59
+ "transformer_blocks.1.ff.linear_out.weight_packed": "model-00001-of-00005.safetensors",
60
+ "transformer_blocks.1.ff.linear_out.weight_scale": "model-00001-of-00005.safetensors",
61
+ "transformer_blocks.1.ff_context.linear_in.weight_packed": "model-00001-of-00005.safetensors",
62
+ "transformer_blocks.1.ff_context.linear_in.weight_scale": "model-00001-of-00005.safetensors",
63
+ "transformer_blocks.1.ff_context.linear_out.weight_packed": "model-00001-of-00005.safetensors",
64
+ "transformer_blocks.1.ff_context.linear_out.weight_scale": "model-00001-of-00005.safetensors",
65
+ "transformer_blocks.2.attn.norm_q.weight": "model-00001-of-00005.safetensors",
66
+ "transformer_blocks.2.attn.norm_k.weight": "model-00001-of-00005.safetensors",
67
+ "transformer_blocks.2.attn.norm_added_q.weight": "model-00001-of-00005.safetensors",
68
+ "transformer_blocks.2.attn.norm_added_k.weight": "model-00001-of-00005.safetensors",
69
+ "transformer_blocks.2.attn.to_q.weight_packed": "model-00001-of-00005.safetensors",
70
+ "transformer_blocks.2.attn.to_q.weight_scale": "model-00001-of-00005.safetensors",
71
+ "transformer_blocks.2.attn.to_k.weight_packed": "model-00001-of-00005.safetensors",
72
+ "transformer_blocks.2.attn.to_k.weight_scale": "model-00001-of-00005.safetensors",
73
+ "transformer_blocks.2.attn.to_v.weight_packed": "model-00001-of-00005.safetensors",
74
+ "transformer_blocks.2.attn.to_v.weight_scale": "model-00001-of-00005.safetensors",
75
+ "transformer_blocks.2.attn.to_out.0.weight_packed": "model-00001-of-00005.safetensors",
76
+ "transformer_blocks.2.attn.to_out.0.weight_scale": "model-00001-of-00005.safetensors",
77
+ "transformer_blocks.2.attn.add_q_proj.weight_packed": "model-00001-of-00005.safetensors",
78
+ "transformer_blocks.2.attn.add_q_proj.weight_scale": "model-00001-of-00005.safetensors",
79
+ "transformer_blocks.2.attn.add_k_proj.weight_packed": "model-00001-of-00005.safetensors",
80
+ "transformer_blocks.2.attn.add_k_proj.weight_scale": "model-00001-of-00005.safetensors",
81
+ "transformer_blocks.2.attn.add_v_proj.weight_packed": "model-00001-of-00005.safetensors",
82
+ "transformer_blocks.2.attn.add_v_proj.weight_scale": "model-00001-of-00005.safetensors",
83
+ "transformer_blocks.2.attn.to_add_out.weight_packed": "model-00001-of-00005.safetensors",
84
+ "transformer_blocks.2.attn.to_add_out.weight_scale": "model-00001-of-00005.safetensors",
85
+ "transformer_blocks.2.ff.linear_in.weight_packed": "model-00001-of-00005.safetensors",
86
+ "transformer_blocks.2.ff.linear_in.weight_scale": "model-00001-of-00005.safetensors",
87
+ "transformer_blocks.2.ff.linear_out.weight_packed": "model-00001-of-00005.safetensors",
88
+ "transformer_blocks.2.ff.linear_out.weight_scale": "model-00001-of-00005.safetensors",
89
+ "transformer_blocks.2.ff_context.linear_in.weight_packed": "model-00001-of-00005.safetensors",
90
+ "transformer_blocks.2.ff_context.linear_in.weight_scale": "model-00001-of-00005.safetensors",
91
+ "transformer_blocks.2.ff_context.linear_out.weight_packed": "model-00001-of-00005.safetensors",
92
+ "transformer_blocks.2.ff_context.linear_out.weight_scale": "model-00001-of-00005.safetensors",
93
+ "transformer_blocks.3.attn.norm_q.weight": "model-00001-of-00005.safetensors",
94
+ "transformer_blocks.3.attn.norm_k.weight": "model-00001-of-00005.safetensors",
95
+ "transformer_blocks.3.attn.norm_added_q.weight": "model-00001-of-00005.safetensors",
96
+ "transformer_blocks.3.attn.norm_added_k.weight": "model-00001-of-00005.safetensors",
97
+ "transformer_blocks.3.attn.to_q.weight_packed": "model-00001-of-00005.safetensors",
98
+ "transformer_blocks.3.attn.to_q.weight_scale": "model-00001-of-00005.safetensors",
99
+ "transformer_blocks.3.attn.to_k.weight_packed": "model-00001-of-00005.safetensors",
100
+ "transformer_blocks.3.attn.to_k.weight_scale": "model-00001-of-00005.safetensors",
101
+ "transformer_blocks.3.attn.to_v.weight_packed": "model-00001-of-00005.safetensors",
102
+ "transformer_blocks.3.attn.to_v.weight_scale": "model-00001-of-00005.safetensors",
103
+ "transformer_blocks.3.attn.to_out.0.weight_packed": "model-00001-of-00005.safetensors",
104
+ "transformer_blocks.3.attn.to_out.0.weight_scale": "model-00001-of-00005.safetensors",
105
+ "transformer_blocks.3.attn.add_q_proj.weight_packed": "model-00001-of-00005.safetensors",
106
+ "transformer_blocks.3.attn.add_q_proj.weight_scale": "model-00001-of-00005.safetensors",
107
+ "transformer_blocks.3.attn.add_k_proj.weight_packed": "model-00001-of-00005.safetensors",
108
+ "transformer_blocks.3.attn.add_k_proj.weight_scale": "model-00001-of-00005.safetensors",
109
+ "transformer_blocks.3.attn.add_v_proj.weight_packed": "model-00001-of-00005.safetensors",
110
+ "transformer_blocks.3.attn.add_v_proj.weight_scale": "model-00001-of-00005.safetensors",
111
+ "transformer_blocks.3.attn.to_add_out.weight_packed": "model-00001-of-00005.safetensors",
112
+ "transformer_blocks.3.attn.to_add_out.weight_scale": "model-00001-of-00005.safetensors",
113
+ "transformer_blocks.3.ff.linear_in.weight_packed": "model-00001-of-00005.safetensors",
114
+ "transformer_blocks.3.ff.linear_in.weight_scale": "model-00001-of-00005.safetensors",
115
+ "transformer_blocks.3.ff.linear_out.weight_packed": "model-00001-of-00005.safetensors",
116
+ "transformer_blocks.3.ff.linear_out.weight_scale": "model-00001-of-00005.safetensors",
117
+ "transformer_blocks.3.ff_context.linear_in.weight_packed": "model-00001-of-00005.safetensors",
118
+ "transformer_blocks.3.ff_context.linear_in.weight_scale": "model-00001-of-00005.safetensors",
119
+ "transformer_blocks.3.ff_context.linear_out.weight_packed": "model-00001-of-00005.safetensors",
120
+ "transformer_blocks.3.ff_context.linear_out.weight_scale": "model-00001-of-00005.safetensors",
121
+ "transformer_blocks.4.attn.norm_q.weight": "model-00001-of-00005.safetensors",
122
+ "transformer_blocks.4.attn.norm_k.weight": "model-00001-of-00005.safetensors",
123
+ "transformer_blocks.4.attn.norm_added_q.weight": "model-00001-of-00005.safetensors",
124
+ "transformer_blocks.4.attn.norm_added_k.weight": "model-00001-of-00005.safetensors",
125
+ "transformer_blocks.4.attn.to_q.weight_packed": "model-00001-of-00005.safetensors",
126
+ "transformer_blocks.4.attn.to_q.weight_scale": "model-00001-of-00005.safetensors",
127
+ "transformer_blocks.4.attn.to_k.weight_packed": "model-00001-of-00005.safetensors",
128
+ "transformer_blocks.4.attn.to_k.weight_scale": "model-00001-of-00005.safetensors",
129
+ "transformer_blocks.4.attn.to_v.weight_packed": "model-00001-of-00005.safetensors",
130
+ "transformer_blocks.4.attn.to_v.weight_scale": "model-00001-of-00005.safetensors",
131
+ "transformer_blocks.4.attn.to_out.0.weight_packed": "model-00001-of-00005.safetensors",
132
+ "transformer_blocks.4.attn.to_out.0.weight_scale": "model-00001-of-00005.safetensors",
133
+ "transformer_blocks.4.attn.add_q_proj.weight_packed": "model-00001-of-00005.safetensors",
134
+ "transformer_blocks.4.attn.add_q_proj.weight_scale": "model-00001-of-00005.safetensors",
135
+ "transformer_blocks.4.attn.add_k_proj.weight_packed": "model-00001-of-00005.safetensors",
136
+ "transformer_blocks.4.attn.add_k_proj.weight_scale": "model-00001-of-00005.safetensors",
137
+ "transformer_blocks.4.attn.add_v_proj.weight_packed": "model-00001-of-00005.safetensors",
138
+ "transformer_blocks.4.attn.add_v_proj.weight_scale": "model-00001-of-00005.safetensors",
139
+ "transformer_blocks.4.attn.to_add_out.weight_packed": "model-00001-of-00005.safetensors",
140
+ "transformer_blocks.4.attn.to_add_out.weight_scale": "model-00001-of-00005.safetensors",
141
+ "transformer_blocks.4.ff.linear_in.weight_packed": "model-00001-of-00005.safetensors",
142
+ "transformer_blocks.4.ff.linear_in.weight_scale": "model-00001-of-00005.safetensors",
143
+ "transformer_blocks.4.ff.linear_out.weight_packed": "model-00001-of-00005.safetensors",
144
+ "transformer_blocks.4.ff.linear_out.weight_scale": "model-00001-of-00005.safetensors",
145
+ "transformer_blocks.4.ff_context.linear_in.weight_packed": "model-00001-of-00005.safetensors",
146
+ "transformer_blocks.4.ff_context.linear_in.weight_scale": "model-00001-of-00005.safetensors",
147
+ "transformer_blocks.4.ff_context.linear_out.weight_packed": "model-00001-of-00005.safetensors",
148
+ "transformer_blocks.4.ff_context.linear_out.weight_scale": "model-00001-of-00005.safetensors",
149
+ "transformer_blocks.5.attn.norm_q.weight": "model-00001-of-00005.safetensors",
150
+ "transformer_blocks.5.attn.norm_k.weight": "model-00001-of-00005.safetensors",
151
+ "transformer_blocks.5.attn.norm_added_q.weight": "model-00001-of-00005.safetensors",
152
+ "transformer_blocks.5.attn.norm_added_k.weight": "model-00001-of-00005.safetensors",
153
+ "transformer_blocks.5.attn.to_q.weight_packed": "model-00001-of-00005.safetensors",
154
+ "transformer_blocks.5.attn.to_q.weight_scale": "model-00001-of-00005.safetensors",
155
+ "transformer_blocks.5.attn.to_k.weight_packed": "model-00001-of-00005.safetensors",
156
+ "transformer_blocks.5.attn.to_k.weight_scale": "model-00001-of-00005.safetensors",
157
+ "transformer_blocks.5.attn.to_v.weight_packed": "model-00001-of-00005.safetensors",
158
+ "transformer_blocks.5.attn.to_v.weight_scale": "model-00001-of-00005.safetensors",
159
+ "transformer_blocks.5.attn.to_out.0.weight_packed": "model-00001-of-00005.safetensors",
160
+ "transformer_blocks.5.attn.to_out.0.weight_scale": "model-00001-of-00005.safetensors",
161
+ "transformer_blocks.5.attn.add_q_proj.weight_packed": "model-00001-of-00005.safetensors",
162
+ "transformer_blocks.5.attn.add_q_proj.weight_scale": "model-00001-of-00005.safetensors",
163
+ "transformer_blocks.5.attn.add_k_proj.weight_packed": "model-00001-of-00005.safetensors",
164
+ "transformer_blocks.5.attn.add_k_proj.weight_scale": "model-00001-of-00005.safetensors",
165
+ "transformer_blocks.5.attn.add_v_proj.weight_packed": "model-00001-of-00005.safetensors",
166
+ "transformer_blocks.5.attn.add_v_proj.weight_scale": "model-00001-of-00005.safetensors",
167
+ "transformer_blocks.5.attn.to_add_out.weight_packed": "model-00001-of-00005.safetensors",
168
+ "transformer_blocks.5.attn.to_add_out.weight_scale": "model-00001-of-00005.safetensors",
169
+ "transformer_blocks.5.ff.linear_in.weight_packed": "model-00001-of-00005.safetensors",
170
+ "transformer_blocks.5.ff.linear_in.weight_scale": "model-00001-of-00005.safetensors",
171
+ "transformer_blocks.5.ff.linear_out.weight_packed": "model-00001-of-00005.safetensors",
172
+ "transformer_blocks.5.ff.linear_out.weight_scale": "model-00001-of-00005.safetensors",
173
+ "transformer_blocks.5.ff_context.linear_in.weight_packed": "model-00001-of-00005.safetensors",
174
+ "transformer_blocks.5.ff_context.linear_in.weight_scale": "model-00001-of-00005.safetensors",
175
+ "transformer_blocks.5.ff_context.linear_out.weight_packed": "model-00001-of-00005.safetensors",
176
+ "transformer_blocks.5.ff_context.linear_out.weight_scale": "model-00001-of-00005.safetensors",
177
+ "transformer_blocks.6.attn.norm_q.weight": "model-00001-of-00005.safetensors",
178
+ "transformer_blocks.6.attn.norm_k.weight": "model-00001-of-00005.safetensors",
179
+ "transformer_blocks.6.attn.norm_added_q.weight": "model-00001-of-00005.safetensors",
180
+ "transformer_blocks.6.attn.norm_added_k.weight": "model-00001-of-00005.safetensors",
181
+ "transformer_blocks.6.attn.to_q.weight_packed": "model-00001-of-00005.safetensors",
182
+ "transformer_blocks.6.attn.to_q.weight_scale": "model-00001-of-00005.safetensors",
183
+ "transformer_blocks.6.attn.to_k.weight_packed": "model-00001-of-00005.safetensors",
184
+ "transformer_blocks.6.attn.to_k.weight_scale": "model-00001-of-00005.safetensors",
185
+ "transformer_blocks.6.attn.to_v.weight_packed": "model-00001-of-00005.safetensors",
186
+ "transformer_blocks.6.attn.to_v.weight_scale": "model-00001-of-00005.safetensors",
187
+ "transformer_blocks.6.attn.to_out.0.weight_packed": "model-00001-of-00005.safetensors",
188
+ "transformer_blocks.6.attn.to_out.0.weight_scale": "model-00001-of-00005.safetensors",
189
+ "transformer_blocks.6.attn.add_q_proj.weight_packed": "model-00001-of-00005.safetensors",
190
+ "transformer_blocks.6.attn.add_q_proj.weight_scale": "model-00001-of-00005.safetensors",
191
+ "transformer_blocks.6.attn.add_k_proj.weight_packed": "model-00001-of-00005.safetensors",
192
+ "transformer_blocks.6.attn.add_k_proj.weight_scale": "model-00001-of-00005.safetensors",
193
+ "transformer_blocks.6.attn.add_v_proj.weight_packed": "model-00001-of-00005.safetensors",
194
+ "transformer_blocks.6.attn.add_v_proj.weight_scale": "model-00001-of-00005.safetensors",
195
+ "transformer_blocks.6.attn.to_add_out.weight_packed": "model-00001-of-00005.safetensors",
196
+ "transformer_blocks.6.attn.to_add_out.weight_scale": "model-00001-of-00005.safetensors",
197
+ "transformer_blocks.6.ff.linear_in.weight_packed": "model-00001-of-00005.safetensors",
198
+ "transformer_blocks.6.ff.linear_in.weight_scale": "model-00001-of-00005.safetensors",
199
+ "transformer_blocks.6.ff.linear_out.weight_packed": "model-00001-of-00005.safetensors",
200
+ "transformer_blocks.6.ff.linear_out.weight_scale": "model-00001-of-00005.safetensors",
201
+ "transformer_blocks.6.ff_context.linear_in.weight_packed": "model-00001-of-00005.safetensors",
202
+ "transformer_blocks.6.ff_context.linear_in.weight_scale": "model-00001-of-00005.safetensors",
203
+ "transformer_blocks.6.ff_context.linear_out.weight_packed": "model-00001-of-00005.safetensors",
204
+ "transformer_blocks.6.ff_context.linear_out.weight_scale": "model-00001-of-00005.safetensors",
205
+ "transformer_blocks.7.attn.norm_q.weight": "model-00001-of-00005.safetensors",
206
+ "transformer_blocks.7.attn.norm_k.weight": "model-00001-of-00005.safetensors",
207
+ "transformer_blocks.7.attn.norm_added_q.weight": "model-00001-of-00005.safetensors",
208
+ "transformer_blocks.7.attn.norm_added_k.weight": "model-00001-of-00005.safetensors",
209
+ "transformer_blocks.7.attn.to_q.weight_packed": "model-00001-of-00005.safetensors",
210
+ "transformer_blocks.7.attn.to_q.weight_scale": "model-00001-of-00005.safetensors",
211
+ "transformer_blocks.7.attn.to_k.weight_packed": "model-00001-of-00005.safetensors",
212
+ "transformer_blocks.7.attn.to_k.weight_scale": "model-00001-of-00005.safetensors",
213
+ "transformer_blocks.7.attn.to_v.weight_packed": "model-00001-of-00005.safetensors",
214
+ "transformer_blocks.7.attn.to_v.weight_scale": "model-00001-of-00005.safetensors",
215
+ "transformer_blocks.7.attn.to_out.0.weight_packed": "model-00001-of-00005.safetensors",
216
+ "transformer_blocks.7.attn.to_out.0.weight_scale": "model-00001-of-00005.safetensors",
217
+ "transformer_blocks.7.attn.add_q_proj.weight_packed": "model-00001-of-00005.safetensors",
218
+ "transformer_blocks.7.attn.add_q_proj.weight_scale": "model-00001-of-00005.safetensors",
219
+ "transformer_blocks.7.attn.add_k_proj.weight_packed": "model-00001-of-00005.safetensors",
220
+ "transformer_blocks.7.attn.add_k_proj.weight_scale": "model-00001-of-00005.safetensors",
221
+ "transformer_blocks.7.attn.add_v_proj.weight_packed": "model-00001-of-00005.safetensors",
222
+ "transformer_blocks.7.attn.add_v_proj.weight_scale": "model-00001-of-00005.safetensors",
223
+ "transformer_blocks.7.attn.to_add_out.weight_packed": "model-00001-of-00005.safetensors",
224
+ "transformer_blocks.7.attn.to_add_out.weight_scale": "model-00001-of-00005.safetensors",
225
+ "transformer_blocks.7.ff.linear_in.weight_packed": "model-00001-of-00005.safetensors",
226
+ "transformer_blocks.7.ff.linear_in.weight_scale": "model-00001-of-00005.safetensors",
227
+ "transformer_blocks.7.ff.linear_out.weight_packed": "model-00001-of-00005.safetensors",
228
+ "transformer_blocks.7.ff.linear_out.weight_scale": "model-00001-of-00005.safetensors",
229
+ "transformer_blocks.7.ff_context.linear_in.weight_packed": "model-00001-of-00005.safetensors",
230
+ "transformer_blocks.7.ff_context.linear_in.weight_scale": "model-00001-of-00005.safetensors",
231
+ "transformer_blocks.7.ff_context.linear_out.weight_packed": "model-00001-of-00005.safetensors",
232
+ "transformer_blocks.7.ff_context.linear_out.weight_scale": "model-00001-of-00005.safetensors",
233
+ "single_transformer_blocks.0.attn.norm_q.weight": "model-00001-of-00005.safetensors",
234
+ "single_transformer_blocks.0.attn.norm_k.weight": "model-00001-of-00005.safetensors",
235
+ "single_transformer_blocks.0.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
236
+ "single_transformer_blocks.0.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
237
+ "single_transformer_blocks.0.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
238
+ "single_transformer_blocks.0.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
239
+ "single_transformer_blocks.1.attn.norm_q.weight": "model-00002-of-00005.safetensors",
240
+ "single_transformer_blocks.1.attn.norm_k.weight": "model-00002-of-00005.safetensors",
241
+ "single_transformer_blocks.1.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
242
+ "single_transformer_blocks.1.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
243
+ "single_transformer_blocks.1.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
244
+ "single_transformer_blocks.1.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
245
+ "single_transformer_blocks.2.attn.norm_q.weight": "model-00002-of-00005.safetensors",
246
+ "single_transformer_blocks.2.attn.norm_k.weight": "model-00002-of-00005.safetensors",
247
+ "single_transformer_blocks.2.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
248
+ "single_transformer_blocks.2.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
249
+ "single_transformer_blocks.2.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
250
+ "single_transformer_blocks.2.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
251
+ "single_transformer_blocks.3.attn.norm_q.weight": "model-00002-of-00005.safetensors",
252
+ "single_transformer_blocks.3.attn.norm_k.weight": "model-00002-of-00005.safetensors",
253
+ "single_transformer_blocks.3.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
254
+ "single_transformer_blocks.3.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
255
+ "single_transformer_blocks.3.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
256
+ "single_transformer_blocks.3.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
257
+ "single_transformer_blocks.4.attn.norm_q.weight": "model-00002-of-00005.safetensors",
258
+ "single_transformer_blocks.4.attn.norm_k.weight": "model-00002-of-00005.safetensors",
259
+ "single_transformer_blocks.4.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
260
+ "single_transformer_blocks.4.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
261
+ "single_transformer_blocks.4.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
262
+ "single_transformer_blocks.4.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
263
+ "single_transformer_blocks.5.attn.norm_q.weight": "model-00002-of-00005.safetensors",
264
+ "single_transformer_blocks.5.attn.norm_k.weight": "model-00002-of-00005.safetensors",
265
+ "single_transformer_blocks.5.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
266
+ "single_transformer_blocks.5.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
267
+ "single_transformer_blocks.5.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
268
+ "single_transformer_blocks.5.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
269
+ "single_transformer_blocks.6.attn.norm_q.weight": "model-00002-of-00005.safetensors",
270
+ "single_transformer_blocks.6.attn.norm_k.weight": "model-00002-of-00005.safetensors",
271
+ "single_transformer_blocks.6.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
272
+ "single_transformer_blocks.6.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
273
+ "single_transformer_blocks.6.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
274
+ "single_transformer_blocks.6.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
275
+ "single_transformer_blocks.7.attn.norm_q.weight": "model-00002-of-00005.safetensors",
276
+ "single_transformer_blocks.7.attn.norm_k.weight": "model-00002-of-00005.safetensors",
277
+ "single_transformer_blocks.7.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
278
+ "single_transformer_blocks.7.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
279
+ "single_transformer_blocks.7.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
280
+ "single_transformer_blocks.7.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
281
+ "single_transformer_blocks.8.attn.norm_q.weight": "model-00002-of-00005.safetensors",
282
+ "single_transformer_blocks.8.attn.norm_k.weight": "model-00002-of-00005.safetensors",
283
+ "single_transformer_blocks.8.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
284
+ "single_transformer_blocks.8.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
285
+ "single_transformer_blocks.8.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
286
+ "single_transformer_blocks.8.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
287
+ "single_transformer_blocks.9.attn.norm_q.weight": "model-00002-of-00005.safetensors",
288
+ "single_transformer_blocks.9.attn.norm_k.weight": "model-00002-of-00005.safetensors",
289
+ "single_transformer_blocks.9.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
290
+ "single_transformer_blocks.9.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
291
+ "single_transformer_blocks.9.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
292
+ "single_transformer_blocks.9.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
293
+ "single_transformer_blocks.10.attn.norm_q.weight": "model-00002-of-00005.safetensors",
294
+ "single_transformer_blocks.10.attn.norm_k.weight": "model-00002-of-00005.safetensors",
295
+ "single_transformer_blocks.10.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
296
+ "single_transformer_blocks.10.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
297
+ "single_transformer_blocks.10.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
298
+ "single_transformer_blocks.10.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
299
+ "single_transformer_blocks.11.attn.norm_q.weight": "model-00002-of-00005.safetensors",
300
+ "single_transformer_blocks.11.attn.norm_k.weight": "model-00002-of-00005.safetensors",
301
+ "single_transformer_blocks.11.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
302
+ "single_transformer_blocks.11.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
303
+ "single_transformer_blocks.11.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
304
+ "single_transformer_blocks.11.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
305
+ "single_transformer_blocks.12.attn.norm_q.weight": "model-00002-of-00005.safetensors",
306
+ "single_transformer_blocks.12.attn.norm_k.weight": "model-00002-of-00005.safetensors",
307
+ "single_transformer_blocks.12.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
308
+ "single_transformer_blocks.12.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
309
+ "single_transformer_blocks.12.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
310
+ "single_transformer_blocks.12.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
311
+ "single_transformer_blocks.13.attn.norm_q.weight": "model-00002-of-00005.safetensors",
312
+ "single_transformer_blocks.13.attn.norm_k.weight": "model-00002-of-00005.safetensors",
313
+ "single_transformer_blocks.13.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
314
+ "single_transformer_blocks.13.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
315
+ "single_transformer_blocks.13.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
316
+ "single_transformer_blocks.13.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
317
+ "single_transformer_blocks.14.attn.norm_q.weight": "model-00002-of-00005.safetensors",
318
+ "single_transformer_blocks.14.attn.norm_k.weight": "model-00002-of-00005.safetensors",
319
+ "single_transformer_blocks.14.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
320
+ "single_transformer_blocks.14.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
321
+ "single_transformer_blocks.14.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
322
+ "single_transformer_blocks.14.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
323
+ "single_transformer_blocks.15.attn.norm_q.weight": "model-00002-of-00005.safetensors",
324
+ "single_transformer_blocks.15.attn.norm_k.weight": "model-00002-of-00005.safetensors",
325
+ "single_transformer_blocks.15.attn.to_qkv_mlp_proj.weight_packed": "model-00002-of-00005.safetensors",
326
+ "single_transformer_blocks.15.attn.to_qkv_mlp_proj.weight_scale": "model-00002-of-00005.safetensors",
327
+ "single_transformer_blocks.15.attn.to_out.weight_packed": "model-00002-of-00005.safetensors",
328
+ "single_transformer_blocks.15.attn.to_out.weight_scale": "model-00002-of-00005.safetensors",
329
+ "single_transformer_blocks.16.attn.norm_q.weight": "model-00002-of-00005.safetensors",
330
+ "single_transformer_blocks.16.attn.norm_k.weight": "model-00002-of-00005.safetensors",
331
+ "single_transformer_blocks.16.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
332
+ "single_transformer_blocks.16.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
333
+ "single_transformer_blocks.16.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
334
+ "single_transformer_blocks.16.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
335
+ "single_transformer_blocks.17.attn.norm_q.weight": "model-00003-of-00005.safetensors",
336
+ "single_transformer_blocks.17.attn.norm_k.weight": "model-00003-of-00005.safetensors",
337
+ "single_transformer_blocks.17.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
338
+ "single_transformer_blocks.17.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
339
+ "single_transformer_blocks.17.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
340
+ "single_transformer_blocks.17.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
341
+ "single_transformer_blocks.18.attn.norm_q.weight": "model-00003-of-00005.safetensors",
342
+ "single_transformer_blocks.18.attn.norm_k.weight": "model-00003-of-00005.safetensors",
343
+ "single_transformer_blocks.18.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
344
+ "single_transformer_blocks.18.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
345
+ "single_transformer_blocks.18.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
346
+ "single_transformer_blocks.18.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
347
+ "single_transformer_blocks.19.attn.norm_q.weight": "model-00003-of-00005.safetensors",
348
+ "single_transformer_blocks.19.attn.norm_k.weight": "model-00003-of-00005.safetensors",
349
+ "single_transformer_blocks.19.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
350
+ "single_transformer_blocks.19.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
351
+ "single_transformer_blocks.19.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
352
+ "single_transformer_blocks.19.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
353
+ "single_transformer_blocks.20.attn.norm_q.weight": "model-00003-of-00005.safetensors",
354
+ "single_transformer_blocks.20.attn.norm_k.weight": "model-00003-of-00005.safetensors",
355
+ "single_transformer_blocks.20.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
356
+ "single_transformer_blocks.20.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
357
+ "single_transformer_blocks.20.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
358
+ "single_transformer_blocks.20.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
359
+ "single_transformer_blocks.21.attn.norm_q.weight": "model-00003-of-00005.safetensors",
360
+ "single_transformer_blocks.21.attn.norm_k.weight": "model-00003-of-00005.safetensors",
361
+ "single_transformer_blocks.21.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
362
+ "single_transformer_blocks.21.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
363
+ "single_transformer_blocks.21.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
364
+ "single_transformer_blocks.21.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
365
+ "single_transformer_blocks.22.attn.norm_q.weight": "model-00003-of-00005.safetensors",
366
+ "single_transformer_blocks.22.attn.norm_k.weight": "model-00003-of-00005.safetensors",
367
+ "single_transformer_blocks.22.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
368
+ "single_transformer_blocks.22.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
369
+ "single_transformer_blocks.22.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
370
+ "single_transformer_blocks.22.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
371
+ "single_transformer_blocks.23.attn.norm_q.weight": "model-00003-of-00005.safetensors",
372
+ "single_transformer_blocks.23.attn.norm_k.weight": "model-00003-of-00005.safetensors",
373
+ "single_transformer_blocks.23.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
374
+ "single_transformer_blocks.23.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
375
+ "single_transformer_blocks.23.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
376
+ "single_transformer_blocks.23.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
377
+ "single_transformer_blocks.24.attn.norm_q.weight": "model-00003-of-00005.safetensors",
378
+ "single_transformer_blocks.24.attn.norm_k.weight": "model-00003-of-00005.safetensors",
379
+ "single_transformer_blocks.24.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
380
+ "single_transformer_blocks.24.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
381
+ "single_transformer_blocks.24.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
382
+ "single_transformer_blocks.24.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
383
+ "single_transformer_blocks.25.attn.norm_q.weight": "model-00003-of-00005.safetensors",
384
+ "single_transformer_blocks.25.attn.norm_k.weight": "model-00003-of-00005.safetensors",
385
+ "single_transformer_blocks.25.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
386
+ "single_transformer_blocks.25.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
387
+ "single_transformer_blocks.25.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
388
+ "single_transformer_blocks.25.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
389
+ "single_transformer_blocks.26.attn.norm_q.weight": "model-00003-of-00005.safetensors",
390
+ "single_transformer_blocks.26.attn.norm_k.weight": "model-00003-of-00005.safetensors",
391
+ "single_transformer_blocks.26.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
392
+ "single_transformer_blocks.26.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
393
+ "single_transformer_blocks.26.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
394
+ "single_transformer_blocks.26.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
395
+ "single_transformer_blocks.27.attn.norm_q.weight": "model-00003-of-00005.safetensors",
396
+ "single_transformer_blocks.27.attn.norm_k.weight": "model-00003-of-00005.safetensors",
397
+ "single_transformer_blocks.27.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
398
+ "single_transformer_blocks.27.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
399
+ "single_transformer_blocks.27.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
400
+ "single_transformer_blocks.27.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
401
+ "single_transformer_blocks.28.attn.norm_q.weight": "model-00003-of-00005.safetensors",
402
+ "single_transformer_blocks.28.attn.norm_k.weight": "model-00003-of-00005.safetensors",
403
+ "single_transformer_blocks.28.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
404
+ "single_transformer_blocks.28.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
405
+ "single_transformer_blocks.28.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
406
+ "single_transformer_blocks.28.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
407
+ "single_transformer_blocks.29.attn.norm_q.weight": "model-00003-of-00005.safetensors",
408
+ "single_transformer_blocks.29.attn.norm_k.weight": "model-00003-of-00005.safetensors",
409
+ "single_transformer_blocks.29.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
410
+ "single_transformer_blocks.29.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
411
+ "single_transformer_blocks.29.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
412
+ "single_transformer_blocks.29.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
413
+ "single_transformer_blocks.30.attn.norm_q.weight": "model-00003-of-00005.safetensors",
414
+ "single_transformer_blocks.30.attn.norm_k.weight": "model-00003-of-00005.safetensors",
415
+ "single_transformer_blocks.30.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
416
+ "single_transformer_blocks.30.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
417
+ "single_transformer_blocks.30.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
418
+ "single_transformer_blocks.30.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
419
+ "single_transformer_blocks.31.attn.norm_q.weight": "model-00003-of-00005.safetensors",
420
+ "single_transformer_blocks.31.attn.norm_k.weight": "model-00003-of-00005.safetensors",
421
+ "single_transformer_blocks.31.attn.to_qkv_mlp_proj.weight_packed": "model-00003-of-00005.safetensors",
422
+ "single_transformer_blocks.31.attn.to_qkv_mlp_proj.weight_scale": "model-00003-of-00005.safetensors",
423
+ "single_transformer_blocks.31.attn.to_out.weight_packed": "model-00003-of-00005.safetensors",
424
+ "single_transformer_blocks.31.attn.to_out.weight_scale": "model-00003-of-00005.safetensors",
425
+ "single_transformer_blocks.32.attn.norm_q.weight": "model-00003-of-00005.safetensors",
426
+ "single_transformer_blocks.32.attn.norm_k.weight": "model-00003-of-00005.safetensors",
427
+ "single_transformer_blocks.32.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
428
+ "single_transformer_blocks.32.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
429
+ "single_transformer_blocks.32.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
430
+ "single_transformer_blocks.32.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
431
+ "single_transformer_blocks.33.attn.norm_q.weight": "model-00004-of-00005.safetensors",
432
+ "single_transformer_blocks.33.attn.norm_k.weight": "model-00004-of-00005.safetensors",
433
+ "single_transformer_blocks.33.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
434
+ "single_transformer_blocks.33.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
435
+ "single_transformer_blocks.33.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
436
+ "single_transformer_blocks.33.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
437
+ "single_transformer_blocks.34.attn.norm_q.weight": "model-00004-of-00005.safetensors",
438
+ "single_transformer_blocks.34.attn.norm_k.weight": "model-00004-of-00005.safetensors",
439
+ "single_transformer_blocks.34.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
440
+ "single_transformer_blocks.34.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
441
+ "single_transformer_blocks.34.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
442
+ "single_transformer_blocks.34.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
443
+ "single_transformer_blocks.35.attn.norm_q.weight": "model-00004-of-00005.safetensors",
444
+ "single_transformer_blocks.35.attn.norm_k.weight": "model-00004-of-00005.safetensors",
445
+ "single_transformer_blocks.35.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
446
+ "single_transformer_blocks.35.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
447
+ "single_transformer_blocks.35.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
448
+ "single_transformer_blocks.35.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
449
+ "single_transformer_blocks.36.attn.norm_q.weight": "model-00004-of-00005.safetensors",
450
+ "single_transformer_blocks.36.attn.norm_k.weight": "model-00004-of-00005.safetensors",
451
+ "single_transformer_blocks.36.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
452
+ "single_transformer_blocks.36.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
453
+ "single_transformer_blocks.36.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
454
+ "single_transformer_blocks.36.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
455
+ "single_transformer_blocks.37.attn.norm_q.weight": "model-00004-of-00005.safetensors",
456
+ "single_transformer_blocks.37.attn.norm_k.weight": "model-00004-of-00005.safetensors",
457
+ "single_transformer_blocks.37.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
458
+ "single_transformer_blocks.37.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
459
+ "single_transformer_blocks.37.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
460
+ "single_transformer_blocks.37.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
461
+ "single_transformer_blocks.38.attn.norm_q.weight": "model-00004-of-00005.safetensors",
462
+ "single_transformer_blocks.38.attn.norm_k.weight": "model-00004-of-00005.safetensors",
463
+ "single_transformer_blocks.38.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
464
+ "single_transformer_blocks.38.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
465
+ "single_transformer_blocks.38.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
466
+ "single_transformer_blocks.38.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
467
+ "single_transformer_blocks.39.attn.norm_q.weight": "model-00004-of-00005.safetensors",
468
+ "single_transformer_blocks.39.attn.norm_k.weight": "model-00004-of-00005.safetensors",
469
+ "single_transformer_blocks.39.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
470
+ "single_transformer_blocks.39.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
471
+ "single_transformer_blocks.39.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
472
+ "single_transformer_blocks.39.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
473
+ "single_transformer_blocks.40.attn.norm_q.weight": "model-00004-of-00005.safetensors",
474
+ "single_transformer_blocks.40.attn.norm_k.weight": "model-00004-of-00005.safetensors",
475
+ "single_transformer_blocks.40.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
476
+ "single_transformer_blocks.40.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
477
+ "single_transformer_blocks.40.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
478
+ "single_transformer_blocks.40.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
479
+ "single_transformer_blocks.41.attn.norm_q.weight": "model-00004-of-00005.safetensors",
480
+ "single_transformer_blocks.41.attn.norm_k.weight": "model-00004-of-00005.safetensors",
481
+ "single_transformer_blocks.41.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
482
+ "single_transformer_blocks.41.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
483
+ "single_transformer_blocks.41.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
484
+ "single_transformer_blocks.41.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
485
+ "single_transformer_blocks.42.attn.norm_q.weight": "model-00004-of-00005.safetensors",
486
+ "single_transformer_blocks.42.attn.norm_k.weight": "model-00004-of-00005.safetensors",
487
+ "single_transformer_blocks.42.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
488
+ "single_transformer_blocks.42.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
489
+ "single_transformer_blocks.42.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
490
+ "single_transformer_blocks.42.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
491
+ "single_transformer_blocks.43.attn.norm_q.weight": "model-00004-of-00005.safetensors",
492
+ "single_transformer_blocks.43.attn.norm_k.weight": "model-00004-of-00005.safetensors",
493
+ "single_transformer_blocks.43.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
494
+ "single_transformer_blocks.43.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
495
+ "single_transformer_blocks.43.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
496
+ "single_transformer_blocks.43.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
497
+ "single_transformer_blocks.44.attn.norm_q.weight": "model-00004-of-00005.safetensors",
498
+ "single_transformer_blocks.44.attn.norm_k.weight": "model-00004-of-00005.safetensors",
499
+ "single_transformer_blocks.44.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
500
+ "single_transformer_blocks.44.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
501
+ "single_transformer_blocks.44.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
502
+ "single_transformer_blocks.44.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
503
+ "single_transformer_blocks.45.attn.norm_q.weight": "model-00004-of-00005.safetensors",
504
+ "single_transformer_blocks.45.attn.norm_k.weight": "model-00004-of-00005.safetensors",
505
+ "single_transformer_blocks.45.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
506
+ "single_transformer_blocks.45.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
507
+ "single_transformer_blocks.45.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
508
+ "single_transformer_blocks.45.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
509
+ "single_transformer_blocks.46.attn.norm_q.weight": "model-00004-of-00005.safetensors",
510
+ "single_transformer_blocks.46.attn.norm_k.weight": "model-00004-of-00005.safetensors",
511
+ "single_transformer_blocks.46.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
512
+ "single_transformer_blocks.46.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
513
+ "single_transformer_blocks.46.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
514
+ "single_transformer_blocks.46.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
515
+ "single_transformer_blocks.47.attn.norm_q.weight": "model-00004-of-00005.safetensors",
516
+ "single_transformer_blocks.47.attn.norm_k.weight": "model-00004-of-00005.safetensors",
517
+ "single_transformer_blocks.47.attn.to_qkv_mlp_proj.weight_packed": "model-00004-of-00005.safetensors",
518
+ "single_transformer_blocks.47.attn.to_qkv_mlp_proj.weight_scale": "model-00004-of-00005.safetensors",
519
+ "single_transformer_blocks.47.attn.to_out.weight_packed": "model-00004-of-00005.safetensors",
520
+ "single_transformer_blocks.47.attn.to_out.weight_scale": "model-00004-of-00005.safetensors",
521
+ "time_guidance_embed.timestep_embedder.linear_1.weight": "model-00004-of-00005.safetensors",
522
+ "time_guidance_embed.timestep_embedder.linear_2.weight": "model-00004-of-00005.safetensors",
523
+ "time_guidance_embed.guidance_embedder.linear_1.weight": "model-00004-of-00005.safetensors",
524
+ "time_guidance_embed.guidance_embedder.linear_2.weight": "model-00005-of-00005.safetensors",
525
+ "double_stream_modulation_img.linear.weight": "model-00005-of-00005.safetensors",
526
+ "double_stream_modulation_txt.linear.weight": "model-00005-of-00005.safetensors",
527
+ "single_stream_modulation.linear.weight": "model-00005-of-00005.safetensors",
528
+ "x_embedder.weight": "model-00005-of-00005.safetensors",
529
+ "context_embedder.weight": "model-00005-of-00005.safetensors",
530
+ "norm_out.linear.weight": "model-00005-of-00005.safetensors",
531
+ "proj_out.weight": "model-00005-of-00005.safetensors"
532
+ }
533
+ }
transformer/quantization_config.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bits": 4,
3
+ "act_bits": 4,
4
+ "data_type": "mx_fp",
5
+ "act_data_type": "mx_fp",
6
+ "group_size": 32,
7
+ "act_group_size": 32,
8
+ "sym": true,
9
+ "act_sym": true,
10
+ "act_dynamic": true,
11
+ "enable_quanted_input": false,
12
+ "static_attention_granularity": "tensor",
13
+ "static_kv_granularity": "tensor",
14
+ "autoround_version": "0.15.0",
15
+ "block_name_to_quantize": "transformer_blocks,single_transformer_blocks",
16
+ "quant_method": "auto-round",
17
+ "packing_format": "auto_round:llm_compressor"
18
+ }
vae/config.json ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "AutoencoderKLFlux2",
3
+ "_diffusers_version": "0.40.0",
4
+ "_name_or_path": "/models/FLUX.2-dev/vae",
5
+ "act_fn": "silu",
6
+ "batch_norm_eps": 0.0001,
7
+ "batch_norm_momentum": 0.1,
8
+ "block_out_channels": [
9
+ 128,
10
+ 256,
11
+ 512,
12
+ 512
13
+ ],
14
+ "decoder_block_out_channels": null,
15
+ "down_block_types": [
16
+ "DownEncoderBlock2D",
17
+ "DownEncoderBlock2D",
18
+ "DownEncoderBlock2D",
19
+ "DownEncoderBlock2D"
20
+ ],
21
+ "force_upcast": true,
22
+ "in_channels": 3,
23
+ "latent_channels": 32,
24
+ "layers_per_block": 2,
25
+ "mid_block_add_attention": true,
26
+ "norm_num_groups": 32,
27
+ "out_channels": 3,
28
+ "patch_size": [
29
+ 2,
30
+ 2
31
+ ],
32
+ "sample_size": 1024,
33
+ "up_block_types": [
34
+ "UpDecoderBlock2D",
35
+ "UpDecoderBlock2D",
36
+ "UpDecoderBlock2D",
37
+ "UpDecoderBlock2D"
38
+ ],
39
+ "use_post_quant_conv": true,
40
+ "use_quant_conv": true
41
+ }
vae/diffusion_pytorch_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca70d2202afe6415bdbcb8793ba8cd99fd159cfe6192381504d6c4d3036e0f04
3
+ size 168120878