Hugging Face's logo

Datasets:
surjo0
/
RAD-Checkpoints

Modalities:
Image
Dataset card Data Studio Files
xet
 Community
1
surjo0 commited on
Commit
dd4b362
·
verified ·
1 Parent(s): 791b91a

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +73 -0
  2. Magick/dlt/run_001.af17baca-b9a1-4a25-869f-92df4ef468e5/training.log +32 -0
  3. Magick/dlt/run_001/checkpoint-2000/optimizer_0/__1_0.distcp +3 -0
  4. Magick/dlt/run_001/checkpoint-2000/pytorch_model_fsdp_0/__1_0.distcp +3 -0
  5. Magick/dlt/run_001/checkpoint-2000/random_states_1.pkl +3 -0
  6. Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/optimizer_0/.metadata +3 -0
  7. Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/optimizer_0/__0_0.distcp +3 -0
  8. Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/pytorch_model_fsdp_0/.metadata +3 -0
  9. Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/pytorch_model_fsdp_0/__0_0.distcp +3 -0
  10. Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/random_states_0.pkl +3 -0
  11. Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/scheduler.bin +3 -0
  12. Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/transformer/config.json +24 -0
  13. Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/training.log +434 -0
  14. Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/validation/collage_step_0001000.png +3 -0
  15. Magick/dlt/run_003.21b84303-af7b-45c5-80da-56d08da757f8/checkpoint-2000/optimizer_0/__2_0.distcp +3 -0
  16. Magick/dlt/run_003.21b84303-af7b-45c5-80da-56d08da757f8/checkpoint-2000/pytorch_model_fsdp_0/__2_0.distcp +3 -0
  17. Magick/dlt/run_003.21b84303-af7b-45c5-80da-56d08da757f8/checkpoint-2000/random_states_2.pkl +3 -0
  18. Magick/dlt/run_006/training.log +32 -0
  19. Magick/dlt/run_007/training.log +32 -0
  20. Magick/dlt/run_010/training.log +32 -0
  21. Magick/dlt/run_016/training.log +14 -0
  22. Magick/dlt/run_019/training.log +14 -0
  23. Magick/dlt/run_023/training.log +232 -0
  24. Magick/dlt/run_026/training.log +233 -0
  25. Magick/flux 2/Baseline/checkpoint-10000/model.safetensors +3 -0
  26. Magick/flux 2/Baseline/checkpoint-10000/optimizer.bin +3 -0
  27. Magick/flux 2/Baseline/checkpoint-10000/random_states_0.pkl +3 -0
  28. Magick/flux 2/Baseline/checkpoint-10000/scheduler.bin +3 -0
  29. Magick/flux 2/Baseline/checkpoint-10000/transformer/config.json +24 -0
  30. Magick/flux 2/Baseline/checkpoint-10000/transformer/diffusion_pytorch_model.safetensors +3 -0
  31. Magick/flux 2/Baseline/checkpoint-8000/model.safetensors +3 -0
  32. Magick/flux 2/Baseline/checkpoint-8000/optimizer.bin +3 -0
  33. Magick/flux 2/Baseline/checkpoint-8000/random_states_0.pkl +3 -0
  34. Magick/flux 2/Baseline/checkpoint-8000/scheduler.bin +3 -0
  35. Magick/flux 2/Baseline/checkpoint-8000/transformer/config.json +24 -0
  36. Magick/flux 2/Baseline/checkpoint-8000/transformer/diffusion_pytorch_model.safetensors +3 -0
  37. Magick/flux 2/Baseline/checkpoint-9000/model.safetensors +3 -0
  38. Magick/flux 2/Baseline/checkpoint-9000/optimizer.bin +3 -0
  39. Magick/flux 2/Baseline/checkpoint-9000/random_states_0.pkl +3 -0
  40. Magick/flux 2/Baseline/checkpoint-9000/scheduler.bin +3 -0
  41. Magick/flux 2/Baseline/checkpoint-9000/transformer/config.json +24 -0
  42. Magick/flux 2/Baseline/checkpoint-9000/transformer/diffusion_pytorch_model.safetensors +3 -0
  43. Magick/flux 2/Baseline/model_index.json +25 -0
  44. Magick/flux 2/Baseline/scheduler/scheduler_config.json +18 -0
  45. Magick/flux 2/Baseline/text_encoder/config.json +71 -0
  46. Magick/flux 2/Baseline/text_encoder/generation_config.json +13 -0
  47. Magick/flux 2/Baseline/text_encoder/model.safetensors +3 -0
  48. Magick/flux 2/Baseline/tokenizer/chat_template.jinja +89 -0
  49. Magick/flux 2/Baseline/tokenizer/tokenizer.json +3 -0
  50. Magick/flux 2/Baseline/tokenizer/tokenizer_config.json +15 -0
.gitattributes CHANGED
@@ -131,3 +131,76 @@ flicker8k/sana[[:space:]]1.6/SpeeD/tokenizer/tokenizer.json filter=lfs diff=lfs
131
  flicker8k/sana[[:space:]]1.6/SRA/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
132
  flicker8k/sana[[:space:]]1.6/RAd/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
133
  flicker8k/sana[[:space:]]1.6/TD/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
131
  flicker8k/sana[[:space:]]1.6/SRA/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
132
  flicker8k/sana[[:space:]]1.6/RAd/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
133
  flicker8k/sana[[:space:]]1.6/TD/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
134
+ Magick/dlt/run_001/checkpoint-2000/optimizer_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
135
+ Magick/dlt/run_001/checkpoint-2000/pytorch_model_fsdp_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
136
+ Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/optimizer_0/.metadata filter=lfs diff=lfs merge=lfs -text
137
+ Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/optimizer_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
138
+ Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/pytorch_model_fsdp_0/.metadata filter=lfs diff=lfs merge=lfs -text
139
+ Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/pytorch_model_fsdp_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
140
+ Magick/dlt/run_003.21b84303-af7b-45c5-80da-56d08da757f8/checkpoint-2000/optimizer_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
141
+ Magick/dlt/run_003.21b84303-af7b-45c5-80da-56d08da757f8/checkpoint-2000/pytorch_model_fsdp_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
142
+ Magick/flux[[:space:]]2/Baseline/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
143
+ Magick/flux[[:space:]]2/RAD/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
144
+ Magick/flux[[:space:]]2/SpeeD/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
145
+ Magick/flux[[:space:]]2/TD/checkpoint-10000/optimizer_0/.metadata filter=lfs diff=lfs merge=lfs -text
146
+ Magick/flux[[:space:]]2/TD/checkpoint-10000/optimizer_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
147
+ Magick/flux[[:space:]]2/TD/checkpoint-10000/pytorch_model_fsdp_0/.metadata filter=lfs diff=lfs merge=lfs -text
148
+ Magick/flux[[:space:]]2/TD/checkpoint-10000/pytorch_model_fsdp_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
149
+ Magick/flux[[:space:]]2/TD/checkpoint-6000/optimizer_0/.metadata filter=lfs diff=lfs merge=lfs -text
150
+ Magick/flux[[:space:]]2/TD/checkpoint-6000/optimizer_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
151
+ Magick/flux[[:space:]]2/TD/checkpoint-6000/pytorch_model_fsdp_0/.metadata filter=lfs diff=lfs merge=lfs -text
152
+ Magick/flux[[:space:]]2/TD/checkpoint-6000/pytorch_model_fsdp_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
153
+ Magick/flux[[:space:]]2/TD/checkpoint-8000/optimizer_0/.metadata filter=lfs diff=lfs merge=lfs -text
154
+ Magick/flux[[:space:]]2/TD/checkpoint-8000/optimizer_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
155
+ Magick/flux[[:space:]]2/TD/checkpoint-8000/pytorch_model_fsdp_0/.metadata filter=lfs diff=lfs merge=lfs -text
156
+ Magick/flux[[:space:]]2/TD/checkpoint-8000/pytorch_model_fsdp_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
157
+ Magick/flux[[:space:]]2/TD/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
158
+ Magick/run_007/checkpoint-2000/optimizer_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
159
+ Magick/run_007/checkpoint-2000/pytorch_model_fsdp_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
160
+ Magick/run_008/checkpoint-2000/optimizer_0/.metadata filter=lfs diff=lfs merge=lfs -text
161
+ Magick/run_008/checkpoint-2000/optimizer_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
162
+ Magick/run_008/checkpoint-2000/pytorch_model_fsdp_0/.metadata filter=lfs diff=lfs merge=lfs -text
163
+ Magick/run_008/checkpoint-2000/pytorch_model_fsdp_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
164
+ Magick/run_009/checkpoint-2000/optimizer_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
165
+ Magick/run_009/checkpoint-2000/pytorch_model_fsdp_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
166
+ Magick/run_010/checkpoint-10000/optimizer_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
167
+ Magick/run_010/checkpoint-10000/pytorch_model_fsdp_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
168
+ Magick/run_010/checkpoint-2000/optimizer_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
169
+ Magick/run_010/checkpoint-2000/pytorch_model_fsdp_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
170
+ Magick/run_010/checkpoint-4000/optimizer_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
171
+ Magick/run_010/checkpoint-4000/pytorch_model_fsdp_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
172
+ Magick/run_010/checkpoint-6000/optimizer_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
173
+ Magick/run_010/checkpoint-6000/pytorch_model_fsdp_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
174
+ Magick/run_010/checkpoint-8000/optimizer_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
175
+ Magick/run_010/checkpoint-8000/pytorch_model_fsdp_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
176
+ Magick/run_012/checkpoint-10000/optimizer_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
177
+ Magick/run_012/checkpoint-10000/pytorch_model_fsdp_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
178
+ Magick/run_012/checkpoint-2000/optimizer_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
179
+ Magick/run_012/checkpoint-2000/pytorch_model_fsdp_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
180
+ Magick/run_012/checkpoint-4000/optimizer_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
181
+ Magick/run_012/checkpoint-4000/pytorch_model_fsdp_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
182
+ Magick/run_012/checkpoint-6000/optimizer_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
183
+ Magick/run_012/checkpoint-6000/pytorch_model_fsdp_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
184
+ Magick/run_012/checkpoint-8000/optimizer_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
185
+ Magick/run_012/checkpoint-8000/pytorch_model_fsdp_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
186
+ ProGamerGov/flux[[:space:]]2/Baseline/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
187
+ ProGamerGov/flux[[:space:]]2/RAD/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
188
+ ProGamerGov/flux[[:space:]]2/SpeeD/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
189
+ ProGamerGov/run_006/checkpoint-2000/optimizer_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
190
+ ProGamerGov/run_006/checkpoint-2000/pytorch_model_fsdp_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
191
+ ProGamerGov/run_006/checkpoint-4000/optimizer_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
192
+ ProGamerGov/run_006/checkpoint-4000/pytorch_model_fsdp_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
193
+ ProGamerGov/run_007/checkpoint-2000/optimizer_0/.metadata filter=lfs diff=lfs merge=lfs -text
194
+ ProGamerGov/run_007/checkpoint-2000/optimizer_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
195
+ ProGamerGov/run_007/checkpoint-2000/pytorch_model_fsdp_0/.metadata filter=lfs diff=lfs merge=lfs -text
196
+ ProGamerGov/run_007/checkpoint-2000/pytorch_model_fsdp_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
197
+ ProGamerGov/run_007/checkpoint-4000/optimizer_0/.metadata filter=lfs diff=lfs merge=lfs -text
198
+ ProGamerGov/run_007/checkpoint-4000/optimizer_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
199
+ ProGamerGov/run_007/checkpoint-4000/pytorch_model_fsdp_0/.metadata filter=lfs diff=lfs merge=lfs -text
200
+ ProGamerGov/run_007/checkpoint-4000/pytorch_model_fsdp_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
201
+ ProGamerGov/run_008/checkpoint-2000/optimizer_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
202
+ ProGamerGov/run_008/checkpoint-2000/pytorch_model_fsdp_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
203
+ ProGamerGov/run_008/checkpoint-4000/optimizer_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
204
+ ProGamerGov/run_008/checkpoint-4000/pytorch_model_fsdp_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
205
+ cub200/sana[[:space:]]0.6/SRA/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
206
+ cub200/sana[[:space:]]0.6/TD/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
Magick/dlt/run_001.af17baca-b9a1-4a25-869f-92df4ef468e5/training.log ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Training started: 2026-08-27 06:01:30
2
+ ==========================================================================================
3
+ Step | Epoch | Loss | Avg100 | LR | Elapsed | ETA
4
+ ------------------------------------------------------------------------------------------
5
+ [06:01:30] Run directory: RESULTS/Magick/run_001
6
+ [06:01:30] Base model: black-forest-labs/FLUX.2-klein-base-4B
7
+ [06:01:30] cuDNN SDPA disabled -- using Flash/math backend for bf16 compatibility
8
+ [06:01:30] Loading pretrained FLUX.2-klein components ...
9
+ [06:01:35] transformer.config.guidance_embeds = False (guidance ignored (base/unguided model))
10
+ [06:01:42] Transformer params: 3,875,544,576 total, 3,875,544,576 trainable
11
+ [06:01:42] VAE scale factor: 8 (x2 for patchify = 16 pixels per packed token)
12
+ [06:01:42] Building dataset ...
13
+ [06:01:42] Step-based training: 10000 steps (~1 epochs over dataset)
14
+ [06:01:54] TD ENABLED: building FSDP-sharded EMA target transformer (ema_tau=0.999, update_freq=1 steps, k_idx=1.0, one_idx=0.3333333333333333, lambda=1.0, weighted=False, sigma_range=[0.002,1.0], rho=7.0, steps=18, sigma_data=1.0)
15
+ [06:01:56]
16
+ [06:01:56] ================================================================
17
+ [06:01:56] Output dir = RESULTS/Magick/run_001
18
+ [06:01:56] Num examples = 74632
19
+ [06:01:56] Num epochs = 1
20
+ [06:01:56] Batch/device = 1
21
+ [06:01:56] Grad accum = 1
22
+ [06:01:56] Effective BS = 3
23
+ [06:01:56] Total optim steps = 10000
24
+ [06:01:56] LR = 1e-05
25
+ [06:01:56] Precision = bf16
26
+ [06:01:56] Text encoder layers= (9, 18, 27)
27
+ [06:01:56] Log every = 10 steps
28
+ [06:01:56] Validate every = 1000 steps
29
+ [06:01:56] Checkpoint every = 2000 steps
30
+ [06:01:56] TD = True
31
+ [06:01:56] ================================================================
32
+ [06:01:56]
Magick/dlt/run_001/checkpoint-2000/optimizer_0/__1_0.distcp ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac831035387577d6bb295acb298df856b6ac03bab086cd830a0e7bcbd3665de2
3
+ size 10335291752
Magick/dlt/run_001/checkpoint-2000/pytorch_model_fsdp_0/__1_0.distcp ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:480601943847e686db9516e035d5f61f028913aaec6c15265cdecb7c7364a374
3
+ size 5167593724
Magick/dlt/run_001/checkpoint-2000/random_states_1.pkl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d1ab6cea96763f8ae2a249779dc678f5d7f0d5d5b1066fdb110ec63c607eb2cf
3
+ size 14864
Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/optimizer_0/.metadata ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a243c1c6230fd848b55256bf5cbf9e49f83913f55c75c9ad600ccb8b1ae7ecc
3
+ size 341700
Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/optimizer_0/__0_0.distcp ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd597863e6b1711e0ca8505fa5eab93f52c4591ac2396e5fae34dfdc2ddd9598
3
+ size 10335188628
Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/pytorch_model_fsdp_0/.metadata ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee014cc81a70146ff86b592f4dcb28cb357e59b9a0f3fd3b1754c7a67979f65e
3
+ size 125590
Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/pytorch_model_fsdp_0/__0_0.distcp ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d678d4dbbaef369b97686ec6a141284dfaa8219ba8c083b11e54704d1b4567a
3
+ size 5167593724
Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/random_states_0.pkl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:23e51b8530532b70cb72de6d8633b04bf45fc33535215502afcc8886fb80e3fc
3
+ size 14864
Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/scheduler.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:823eea85e63f13643305fc1c2787fec8eae93dbe2abe3290cf7ca3a769e0874f
3
+ size 1000
Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/checkpoint-2000/transformer/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "Flux2Transformer2DModel",
3
+ "_diffusers_version": "0.37.1",
4
+ "_name_or_path": "black-forest-labs/FLUX.2-klein-base-4B",
5
+ "attention_head_dim": 128,
6
+ "axes_dims_rope": [
7
+ 32,
8
+ 32,
9
+ 32,
10
+ 32
11
+ ],
12
+ "eps": 1e-06,
13
+ "guidance_embeds": false,
14
+ "in_channels": 128,
15
+ "joint_attention_dim": 7680,
16
+ "mlp_ratio": 3.0,
17
+ "num_attention_heads": 24,
18
+ "num_layers": 5,
19
+ "num_single_layers": 20,
20
+ "out_channels": null,
21
+ "patch_size": 1,
22
+ "rope_theta": 2000,
23
+ "timestep_guidance_channels": 256
24
+ }
Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/training.log ADDED
@@ -0,0 +1,434 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Training started: 2026-08-27 03:22:11
2
+ ==========================================================================================
3
+ Step | Epoch | Loss | Avg100 | LR | Elapsed | ETA
4
+ ------------------------------------------------------------------------------------------
5
+ [03:22:11] Run directory: RESULTS/Magick/run_002
6
+ [03:22:11] Base model: black-forest-labs/FLUX.2-klein-base-4B
7
+ [03:22:11] cuDNN SDPA disabled -- using Flash/math backend for bf16 compatibility
8
+ [03:22:11] Loading pretrained FLUX.2-klein components ...
9
+ [03:22:16] transformer.config.guidance_embeds = False (guidance ignored (base/unguided model))
10
+ [03:22:22] Transformer params: 3,875,544,576 total, 3,875,544,576 trainable
11
+ [03:22:22] VAE scale factor: 8 (x2 for patchify = 16 pixels per packed token)
12
+ [03:22:22] Building dataset ...
13
+ [03:22:22] Step-based training: 10000 steps (~1 epochs over dataset)
14
+ [03:22:33] TD ENABLED: building FSDP-sharded EMA target transformer (ema_tau=0.999, update_freq=1 steps, k_idx=1.0, one_idx=0.3333333333333333, lambda=1.0, weighted=False, sigma_range=[0.002,1.0], rho=7.0, steps=18, sigma_data=1.0)
15
+ [03:22:35]
16
+ [03:22:35] ================================================================
17
+ [03:22:35] Output dir = RESULTS/Magick/run_002
18
+ [03:22:35] Num examples = 74632
19
+ [03:22:35] Num epochs = 1
20
+ [03:22:35] Batch/device = 1
21
+ [03:22:35] Grad accum = 1
22
+ [03:22:35] Effective BS = 3
23
+ [03:22:35] Total optim steps = 10000
24
+ [03:22:35] LR = 1e-05
25
+ [03:22:35] Precision = bf16
26
+ [03:22:35] Text encoder layers= (9, 18, 27)
27
+ [03:22:35] Log every = 10 steps
28
+ [03:22:35] Validate every = 1000 steps
29
+ [03:22:35] Checkpoint every = 2000 steps
30
+ [03:22:35] TD = True
31
+ [03:22:35] ================================================================
32
+ [03:22:35]
33
+ 10 | 0 | 1.327983 | 1.327983 | 1.00e-06 | 0:01:00 | 16:53:52
34
+ [03:23:12] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000077
35
+ 20 | 0 | 1.239815 | 1.283899 | 2.00e-06 | 0:01:37 | 13:27:08
36
+ [03:23:48] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000089
37
+ 30 | 0 | 1.111842 | 1.226546 | 3.00e-06 | 0:02:14 | 12:25:52
38
+ [03:24:26] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000053
39
+ 40 | 0 | 0.718584 | 1.099556 | 4.00e-06 | 0:02:52 | 11:56:18
40
+ [03:25:04] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000188
41
+ 50 | 0 | 0.529301 | 0.985505 | 5.00e-06 | 0:03:30 | 11:37:21
42
+ [03:25:41] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000254
43
+ 60 | 0 | 0.801975 | 0.954916 | 6.00e-06 | 0:04:07 | 11:23:43
44
+ [03:26:19] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000138
45
+ 70 | 0 | 0.542700 | 0.896028 | 7.00e-06 | 0:04:44 | 11:12:42
46
+ [03:26:56] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000187
47
+ 80 | 0 | 0.522081 | 0.849285 | 8.00e-06 | 0:05:22 | 11:06:29
48
+ [03:27:34] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000157
49
+ 90 | 0 | 0.587526 | 0.820200 | 9.00e-06 | 0:05:59 | 11:00:24
50
+ [03:28:11] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000166
51
+ 100 | 0 | 1.040526 | 0.842233 | 1.00e-05 | 0:06:37 | 10:55:31
52
+ [03:28:48] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000108
53
+ 110 | 0 | 0.860861 | 0.843926 | 1.00e-05 | 0:07:15 | 10:52:01
54
+ [03:29:26] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000163
55
+ 120 | 0 | 0.564158 | 0.820612 | 1.00e-05 | 0:07:52 | 10:48:45
56
+ [03:30:04] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000141
57
+ 130 | 0 | 0.476915 | 0.794174 | 1.00e-05 | 0:08:30 | 10:45:56
58
+ [03:30:42] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000399
59
+ 140 | 0 | 1.032545 | 0.811201 | 1.00e-05 | 0:09:08 | 10:43:19
60
+ [03:31:19] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000107
61
+ 150 | 0 | 1.079600 | 0.829094 | 1.00e-05 | 0:09:45 | 10:40:49
62
+ [03:31:57] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000092
63
+ 160 | 0 | 1.141017 | 0.848589 | 1.00e-05 | 0:10:23 | 10:38:43
64
+ [03:32:34] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000010
65
+ 170 | 0 | 0.517098 | 0.829090 | 1.00e-05 | 0:11:00 | 10:36:39
66
+ [03:33:12] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000782
67
+ 180 | 0 | 0.732383 | 0.823717 | 1.00e-05 | 0:11:38 | 10:34:59
68
+ [03:33:49] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000126
69
+ 190 | 0 | 0.473995 | 0.805311 | 1.00e-05 | 0:12:15 | 10:33:20
70
+ [03:34:27] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000164
71
+ 200 | 0 | 0.537342 | 0.791912 | 1.00e-05 | 0:12:52 | 10:31:15
72
+ [03:35:04] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000110
73
+ 210 | 0 | 0.677194 | 0.786449 | 1.00e-05 | 0:13:30 | 10:29:51
74
+ [03:35:42] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
75
+ 220 | 0 | 0.483012 | 0.772657 | 1.00e-05 | 0:14:08 | 10:28:43
76
+ [03:36:20] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000652
77
+ 230 | 0 | 0.513717 | 0.761399 | 1.00e-05 | 0:14:46 | 10:27:24
78
+ [03:36:57] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000190
79
+ 240 | 0 | 0.950876 | 0.769293 | 1.00e-05 | 0:15:23 | 10:26:06
80
+ [03:37:35] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000167
81
+ 250 | 0 | 0.861754 | 0.772992 | 9.99e-06 | 0:16:01 | 10:24:50
82
+ [03:38:12] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000121
83
+ 260 | 0 | 0.624339 | 0.767274 | 9.99e-06 | 0:16:38 | 10:23:43
84
+ [03:38:50] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000135
85
+ 270 | 0 | 0.654978 | 0.763115 | 9.99e-06 | 0:17:16 | 10:22:28
86
+ [03:39:27] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000141
87
+ 280 | 0 | 0.589645 | 0.756920 | 9.99e-06 | 0:17:53 | 10:21:06
88
+ [03:40:05] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000179
89
+ 290 | 0 | 0.881581 | 0.761219 | 9.99e-06 | 0:18:30 | 10:19:58
90
+ [03:40:42] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000149
91
+ 300 | 0 | 0.542774 | 0.753937 | 9.99e-06 | 0:19:08 | 10:18:57
92
+ [03:41:20] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000103
93
+ 310 | 0 | 0.787479 | 0.755019 | 9.99e-06 | 0:19:46 | 10:17:59
94
+ [03:41:57] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000151
95
+ 320 | 0 | 0.472911 | 0.746203 | 9.99e-06 | 0:20:23 | 10:16:57
96
+ [03:42:35] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000170
97
+ 330 | 0 | 0.768495 | 0.746879 | 9.99e-06 | 0:21:01 | 10:16:06
98
+ [03:43:13] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000196
99
+ 340 | 0 | 0.487723 | 0.739257 | 9.99e-06 | 0:21:39 | 10:15:15
100
+ [03:43:50] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000266
101
+ 350 | 0 | 0.832604 | 0.741924 | 9.98e-06 | 0:22:16 | 10:14:18
102
+ [03:44:28] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000169
103
+ 360 | 0 | 0.625798 | 0.738698 | 9.98e-06 | 0:22:54 | 10:13:25
104
+ [03:45:06] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000123
105
+ 370 | 0 | 0.936838 | 0.744053 | 9.98e-06 | 0:23:31 | 10:12:29
106
+ [03:45:43] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000169
107
+ 380 | 0 | 0.767905 | 0.744681 | 9.98e-06 | 0:24:09 | 10:11:37
108
+ [03:46:21] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000158
109
+ 390 | 0 | 0.703557 | 0.743626 | 9.98e-06 | 0:24:46 | 10:10:36
110
+ [03:46:58] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000136
111
+ 400 | 0 | 1.073461 | 0.751872 | 9.98e-06 | 0:25:24 | 10:09:47
112
+ [03:47:36] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000053
113
+ 410 | 0 | 0.597282 | 0.748102 | 9.98e-06 | 0:26:01 | 10:08:51
114
+ [03:48:13] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000127
115
+ 420 | 0 | 0.941752 | 0.752712 | 9.97e-06 | 0:26:39 | 10:07:59
116
+ [03:48:50] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000104
117
+ 430 | 0 | 0.475320 | 0.746261 | 9.97e-06 | 0:27:16 | 10:07:10
118
+ [03:49:28] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000174
119
+ 440 | 0 | 0.464786 | 0.739864 | 9.97e-06 | 0:27:54 | 10:06:19
120
+ [03:50:05] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000288
121
+ 450 | 0 | 0.544818 | 0.735530 | 9.97e-06 | 0:28:31 | 10:05:18
122
+ [03:50:42] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000154
123
+ 460 | 0 | 0.512676 | 0.730685 | 9.97e-06 | 0:29:09 | 10:04:38
124
+ [03:51:20] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000239
125
+ 470 | 0 | 0.579516 | 0.727469 | 9.97e-06 | 0:29:47 | 10:03:59
126
+ [03:51:58] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000155
127
+ 480 | 0 | 0.797770 | 0.728933 | 9.96e-06 | 0:30:24 | 10:03:05
128
+ [03:52:36] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000160
129
+ 490 | 0 | 0.828944 | 0.730974 | 9.96e-06 | 0:31:01 | 10:01:59
130
+ [03:53:12] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000161
131
+ 500 | 0 | 0.486766 | 0.726090 | 9.96e-06 | 0:31:37 | 10:00:50
132
+ [03:53:48] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000564
133
+ 510 | 0 | 1.142440 | 0.734254 | 9.96e-06 | 0:32:15 | 10:00:13
134
+ [03:54:26] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000071
135
+ 520 | 0 | 0.645079 | 0.732539 | 9.96e-06 | 0:32:53 | 9:59:31
136
+ [03:55:04] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000115
137
+ 530 | 0 | 0.495293 | 0.728063 | 9.95e-06 | 0:33:30 | 9:58:48
138
+ [03:55:42] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000129
139
+ 540 | 0 | 0.584426 | 0.725403 | 9.95e-06 | 0:34:08 | 9:58:11
140
+ [03:56:20] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000163
141
+ 550 | 0 | 0.991824 | 0.730247 | 9.95e-06 | 0:34:46 | 9:57:31
142
+ [03:56:58] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000104
143
+ 560 | 0 | 0.661446 | 0.729018 | 9.95e-06 | 0:35:24 | 9:56:46
144
+ [03:57:35] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000178
145
+ 570 | 0 | 0.483156 | 0.724705 | 9.94e-06 | 0:36:01 | 9:56:02
146
+ [03:58:13] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000566
147
+ 580 | 0 | 0.889484 | 0.727546 | 9.94e-06 | 0:36:39 | 9:55:26
148
+ [03:58:51] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000148
149
+ 590 | 0 | 1.081373 | 0.733543 | 9.94e-06 | 0:37:17 | 9:54:40
150
+ [03:59:28] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000100
151
+ 600 | 0 | 0.702017 | 0.733018 | 9.94e-06 | 0:37:54 | 9:53:54
152
+ [04:00:06] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000146
153
+ 610 | 0 | 0.781813 | 0.733817 | 9.93e-06 | 0:38:32 | 9:53:11
154
+ [04:00:43] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000206
155
+ 620 | 0 | 0.511101 | 0.730225 | 9.93e-06 | 0:39:10 | 9:52:34
156
+ [04:01:21] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000191
157
+ 630 | 0 | 0.631750 | 0.728662 | 9.93e-06 | 0:39:47 | 9:51:54
158
+ [04:01:59] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000133
159
+ 640 | 0 | 1.049259 | 0.733672 | 9.93e-06 | 0:40:25 | 9:51:13
160
+ [04:02:37] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000116
161
+ 650 | 0 | 0.853146 | 0.735510 | 9.92e-06 | 0:41:03 | 9:50:31
162
+ [04:03:14] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000148
163
+ 660 | 0 | 1.066422 | 0.740523 | 9.92e-06 | 0:41:39 | 9:49:34
164
+ [04:03:51] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000098
165
+ 670 | 0 | 0.946678 | 0.743600 | 9.92e-06 | 0:42:17 | 9:48:49
166
+ [04:04:28] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000167
167
+ 680 | 0 | 0.915616 | 0.746130 | 9.92e-06 | 0:42:54 | 9:48:08
168
+ [04:05:06] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
169
+ 690 | 0 | 0.579759 | 0.743719 | 9.91e-06 | 0:43:32 | 9:47:32
170
+ [04:05:44] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000127
171
+ 700 | 0 | 0.555281 | 0.741027 | 9.91e-06 | 0:44:10 | 9:46:49
172
+ [04:06:21] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000134
173
+ 710 | 0 | 0.562942 | 0.738519 | 9.91e-06 | 0:44:47 | 9:46:09
174
+ [04:06:59] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000145
175
+ 720 | 0 | 0.552789 | 0.735939 | 9.90e-06 | 0:45:26 | 9:45:35
176
+ [04:07:37] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000138
177
+ 730 | 0 | 0.888568 | 0.738030 | 9.90e-06 | 0:46:02 | 9:44:42
178
+ [04:08:14] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000182
179
+ 740 | 0 | 0.640545 | 0.736712 | 9.90e-06 | 0:46:40 | 9:43:58
180
+ [04:08:51] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000153
181
+ 750 | 0 | 1.074535 | 0.741217 | 9.89e-06 | 0:47:17 | 9:43:16
182
+ [04:09:29] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000113
183
+ 760 | 0 | 0.530110 | 0.738439 | 9.89e-06 | 0:47:55 | 9:42:39
184
+ [04:10:07] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000149
185
+ 770 | 0 | 0.560525 | 0.736128 | 9.89e-06 | 0:48:33 | 9:42:02
186
+ [04:10:44] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000139
187
+ 780 | 0 | 1.083861 | 0.740587 | 9.88e-06 | 0:49:10 | 9:41:15
188
+ [04:11:21] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000123
189
+ 790 | 0 | 0.896892 | 0.742565 | 9.88e-06 | 0:49:46 | 9:40:16
190
+ [04:11:58] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000189
191
+ 800 | 0 | 0.896386 | 0.744488 | 9.88e-06 | 0:50:23 | 9:39:33
192
+ [04:12:35] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000148
193
+ 810 | 0 | 0.680541 | 0.743698 | 9.87e-06 | 0:51:00 | 9:38:45
194
+ [04:13:12] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000148
195
+ 820 | 0 | 1.153964 | 0.748702 | 9.87e-06 | 0:51:38 | 9:38:05
196
+ [04:13:49] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000087
197
+ 830 | 0 | 0.676096 | 0.747827 | 9.87e-06 | 0:52:15 | 9:37:20
198
+ [04:14:26] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000164
199
+ 840 | 0 | 0.466228 | 0.744475 | 9.86e-06 | 0:52:51 | 9:36:24
200
+ [04:15:03] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000562
201
+ 850 | 0 | 1.154112 | 0.749294 | 9.86e-06 | 0:53:28 | 9:35:33
202
+ [04:15:39] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000039
203
+ 860 | 0 | 1.037642 | 0.752647 | 9.86e-06 | 0:54:05 | 9:34:49
204
+ [04:16:16] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000147
205
+ 870 | 0 | 0.624911 | 0.751178 | 9.85e-06 | 0:54:42 | 9:34:05
206
+ [04:16:53] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000137
207
+ 880 | 0 | 0.845399 | 0.752249 | 9.85e-06 | 0:55:19 | 9:33:20
208
+ [04:17:30] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000141
209
+ 890 | 0 | 1.041259 | 0.755496 | 9.84e-06 | 0:55:57 | 9:32:44
210
+ [04:18:08] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000114
211
+ 900 | 0 | 0.500237 | 0.752660 | 9.84e-06 | 0:56:35 | 9:32:08
212
+ [04:18:46] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000156
213
+ 910 | 0 | 0.455834 | 0.749398 | 9.84e-06 | 0:57:12 | 9:31:29
214
+ [04:19:24] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000323
215
+ 920 | 0 | 0.568435 | 0.747431 | 9.83e-06 | 0:57:50 | 9:30:50
216
+ [04:20:01] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000147
217
+ 930 | 0 | 1.139783 | 0.751650 | 9.83e-06 | 0:58:27 | 9:30:08
218
+ [04:20:39] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000076
219
+ 940 | 0 | 1.128572 | 0.755660 | 9.82e-06 | 0:59:05 | 9:29:31
220
+ [04:21:16] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000009
221
+ 950 | 0 | 0.643331 | 0.754478 | 9.82e-06 | 0:59:43 | 9:28:53
222
+ [04:21:54] TD diagnostics: frac_valid=1.000, td_loss_mean=0.001499
223
+ 960 | 0 | 0.641123 | 0.753297 | 9.81e-06 | 1:00:20 | 9:28:13
224
+ [04:22:32] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000133
225
+ 970 | 0 | 1.132375 | 0.757205 | 9.81e-06 | 1:00:58 | 9:27:36
226
+ [04:23:09] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000082
227
+ 980 | 0 | 0.564641 | 0.755240 | 9.81e-06 | 1:01:36 | 9:27:00
228
+ [04:23:47] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000126
229
+ 990 | 0 | 1.149226 | 0.759220 | 9.80e-06 | 1:02:13 | 9:26:19
230
+ [04:24:25] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000053
231
+ 1000 | 0 | 1.137289 | 0.763000 | 9.80e-06 | 1:02:51 | 9:25:42
232
+ [04:25:02] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000063
233
+ [04:25:27] Generating 16 validation images ...
234
+ [04:31:56] Validation step 1000: 16 images -> RESULTS/Magick/run_002/validation/collage_step_0001000.png
235
+ 1010 | 0 | 0.888276 | 0.758603 | 9.79e-06 | 1:10:21 | 10:26:16
236
+ [04:32:33] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000135
237
+ 1020 | 0 | 1.128024 | 0.757485 | 9.79e-06 | 1:10:58 | 10:24:55
238
+ [04:33:10] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000010
239
+ 1030 | 0 | 0.540924 | 0.751776 | 9.78e-06 | 1:11:36 | 10:23:36
240
+ [04:33:47] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000127
241
+ 1040 | 0 | 0.553961 | 0.750130 | 9.78e-06 | 1:12:14 | 10:22:21
242
+ [04:34:25] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000157
243
+ 1050 | 0 | 0.921647 | 0.754053 | 9.77e-06 | 1:12:51 | 10:21:01
244
+ [04:35:03] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000183
245
+ 1060 | 0 | 1.088598 | 0.756920 | 9.77e-06 | 1:13:29 | 10:19:50
246
+ [04:35:41] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000108
247
+ 1070 | 0 | 0.586962 | 0.757362 | 9.76e-06 | 1:14:06 | 10:18:28
248
+ [04:36:17] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000227
249
+ 1080 | 0 | 0.802915 | 0.760171 | 9.76e-06 | 1:14:42 | 10:17:03
250
+ [04:36:54] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000156
251
+ 1090 | 0 | 0.456984 | 0.758865 | 9.76e-06 | 1:15:19 | 10:15:44
252
+ [04:37:31] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000175
253
+ 1100 | 0 | 0.847960 | 0.756940 | 9.75e-06 | 1:15:57 | 10:14:32
254
+ [04:38:08] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000132
255
+ 1110 | 0 | 0.427861 | 0.752610 | 9.75e-06 | 1:16:35 | 10:13:21
256
+ [04:38:46] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000212
257
+ 1120 | 0 | 0.935933 | 0.756327 | 9.74e-06 | 1:17:12 | 10:12:10
258
+ [04:39:24] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000142
259
+ 1130 | 0 | 0.602479 | 0.757583 | 9.74e-06 | 1:17:50 | 10:11:00
260
+ [04:40:01] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000136
261
+ 1140 | 0 | 0.649681 | 0.753754 | 9.73e-06 | 1:18:28 | 10:09:52
262
+ [04:40:39] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000132
263
+ 1150 | 0 | 0.569779 | 0.748656 | 9.73e-06 | 1:19:06 | 10:08:43
264
+ [04:41:17] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000159
265
+ 1160 | 0 | 1.138126 | 0.748627 | 9.72e-06 | 1:19:43 | 10:07:32
266
+ [04:41:54] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000014
267
+ 1170 | 0 | 0.815637 | 0.751613 | 9.71e-06 | 1:20:21 | 10:06:26
268
+ [04:42:32] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000153
269
+ 1180 | 0 | 0.974680 | 0.754036 | 9.71e-06 | 1:20:59 | 10:05:21
270
+ [04:43:10] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000201
271
+ 1190 | 0 | 0.702098 | 0.756317 | 9.70e-06 | 1:21:36 | 10:04:13
272
+ [04:43:48] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000165
273
+ 1200 | 0 | 0.595835 | 0.756901 | 9.70e-06 | 1:22:13 | 10:02:56
274
+ [04:44:24] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000141
275
+ 1210 | 0 | 0.468061 | 0.754810 | 9.69e-06 | 1:22:49 | 10:01:38
276
+ [04:45:00] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000179
277
+ 1220 | 0 | 1.014505 | 0.760125 | 9.69e-06 | 1:23:26 | 10:00:30
278
+ [04:45:38] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000138
279
+ 1230 | 0 | 1.136581 | 0.766354 | 9.68e-06 | 1:24:02 | 9:59:15
280
+ [04:46:14] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000050
281
+ 1240 | 0 | 0.621056 | 0.763055 | 9.68e-06 | 1:24:40 | 9:58:08
282
+ [04:46:51] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000170
283
+ 1250 | 0 | 0.935706 | 0.763795 | 9.67e-06 | 1:25:18 | 9:57:06
284
+ [04:47:29] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000162
285
+ 1260 | 0 | 0.539614 | 0.762948 | 9.67e-06 | 1:25:55 | 9:56:01
286
+ [04:48:07] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000144
287
+ 1270 | 0 | 1.115642 | 0.767554 | 9.66e-06 | 1:26:33 | 9:54:59
288
+ [04:48:45] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000083
289
+ 1280 | 0 | 0.692885 | 0.768587 | 9.65e-06 | 1:27:10 | 9:53:55
290
+ [04:49:22] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
291
+ 1290 | 0 | 0.494580 | 0.764717 | 9.65e-06 | 1:27:48 | 9:52:53
292
+ [04:50:00] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000221
293
+ 1300 | 0 | 0.501018 | 0.764299 | 9.64e-06 | 1:28:25 | 9:51:43
294
+ [04:50:36] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000160
295
+ 1310 | 0 | 0.991786 | 0.766342 | 9.64e-06 | 1:29:02 | 9:50:41
296
+ [04:51:14] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000153
297
+ 1320 | 0 | 0.988064 | 0.771494 | 9.63e-06 | 1:29:40 | 9:49:40
298
+ [04:51:52] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000151
299
+ 1330 | 0 | 1.146335 | 0.775272 | 9.62e-06 | 1:30:17 | 9:48:37
300
+ [04:52:29] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000064
301
+ 1340 | 0 | 0.478994 | 0.775185 | 9.62e-06 | 1:30:55 | 9:47:34
302
+ [04:53:06] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000155
303
+ 1350 | 0 | 0.636829 | 0.773227 | 9.61e-06 | 1:31:32 | 9:46:33
304
+ [04:53:44] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
305
+ 1360 | 0 | 0.695968 | 0.773929 | 9.61e-06 | 1:32:10 | 9:45:35
306
+ [04:54:22] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000153
307
+ 1370 | 0 | 0.984127 | 0.774402 | 9.60e-06 | 1:32:48 | 9:44:35
308
+ [04:54:59] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000148
309
+ 1380 | 0 | 0.498092 | 0.771704 | 9.59e-06 | 1:33:25 | 9:43:31
310
+ [04:55:36] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000332
311
+ 1390 | 0 | 0.919502 | 0.773863 | 9.59e-06 | 1:34:01 | 9:42:26
312
+ [04:56:13] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000131
313
+ 1400 | 0 | 0.902781 | 0.772156 | 9.58e-06 | 1:34:39 | 9:41:30
314
+ [04:56:51] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000182
315
+ 1410 | 0 | 0.717631 | 0.773360 | 9.57e-06 | 1:35:17 | 9:40:31
316
+ [04:57:28] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000174
317
+ 1420 | 0 | 0.805373 | 0.771996 | 9.57e-06 | 1:35:54 | 9:39:31
318
+ [04:58:06] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000169
319
+ 1430 | 0 | 1.147955 | 0.778722 | 9.56e-06 | 1:36:32 | 9:38:32
320
+ [04:58:43] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000015
321
+ 1440 | 0 | 0.976684 | 0.783841 | 9.55e-06 | 1:37:09 | 9:37:35
322
+ [04:59:21] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000133
323
+ 1450 | 0 | 1.030757 | 0.788701 | 9.55e-06 | 1:37:47 | 9:36:38
324
+ [04:59:59] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000153
325
+ 1460 | 0 | 0.497547 | 0.788549 | 9.54e-06 | 1:38:25 | 9:35:40
326
+ [05:00:36] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000247
327
+ 1470 | 0 | 1.064664 | 0.793401 | 9.53e-06 | 1:39:02 | 9:34:44
328
+ [05:01:14] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000124
329
+ 1480 | 0 | 0.618828 | 0.791612 | 9.53e-06 | 1:39:40 | 9:33:46
330
+ [05:01:51] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000134
331
+ 1490 | 0 | 0.741809 | 0.790740 | 9.52e-06 | 1:40:17 | 9:32:51
332
+ [05:02:29] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000148
333
+ 1500 | 0 | 0.894974 | 0.794822 | 9.51e-06 | 1:40:55 | 9:31:54
334
+ [05:03:07] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
335
+ 1510 | 0 | 0.706496 | 0.790463 | 9.51e-06 | 1:41:33 | 9:30:58
336
+ [05:03:44] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
337
+ 1520 | 0 | 0.986064 | 0.793873 | 9.50e-06 | 1:42:10 | 9:30:03
338
+ [05:04:22] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000148
339
+ 1530 | 0 | 0.504733 | 0.793967 | 9.49e-06 | 1:42:48 | 9:29:06
340
+ [05:04:59] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000159
341
+ 1540 | 0 | 1.159692 | 0.799720 | 9.49e-06 | 1:43:26 | 9:28:13
342
+ [05:05:37] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000048
343
+ 1550 | 0 | 0.785022 | 0.797652 | 9.48e-06 | 1:44:03 | 9:27:14
344
+ [05:06:14] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000165
345
+ 1560 | 0 | 0.574779 | 0.796785 | 9.47e-06 | 1:44:39 | 9:26:15
346
+ [05:06:51] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000130
347
+ 1570 | 0 | 0.843896 | 0.800392 | 9.47e-06 | 1:45:16 | 9:25:16
348
+ [05:07:28] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000164
349
+ 1580 | 0 | 0.767338 | 0.799171 | 9.46e-06 | 1:45:54 | 9:24:23
350
+ [05:08:05] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000203
351
+ 1590 | 0 | 0.831158 | 0.796669 | 9.45e-06 | 1:46:31 | 9:23:28
352
+ [05:08:43] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000206
353
+ 1600 | 0 | 0.640408 | 0.796053 | 9.44e-06 | 1:47:09 | 9:22:35
354
+ [05:09:21] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000156
355
+ 1610 | 0 | 1.153387 | 0.799768 | 9.44e-06 | 1:47:47 | 9:21:43
356
+ [05:09:59] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000021
357
+ 1620 | 0 | 0.521520 | 0.799873 | 9.43e-06 | 1:48:24 | 9:20:48
358
+ [05:10:36] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000126
359
+ 1630 | 0 | 0.450446 | 0.798060 | 9.42e-06 | 1:49:02 | 9:19:55
360
+ [05:11:14] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000174
361
+ 1640 | 0 | 0.506300 | 0.792630 | 9.41e-06 | 1:49:40 | 9:19:02
362
+ [05:11:51] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000170
363
+ 1650 | 0 | 0.931115 | 0.793410 | 9.41e-06 | 1:50:17 | 9:18:10
364
+ [05:12:29] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000118
365
+ 1660 | 0 | 0.793730 | 0.790683 | 9.40e-06 | 1:50:55 | 9:17:17
366
+ [05:13:06] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000135
367
+ 1670 | 0 | 0.704635 | 0.788262 | 9.39e-06 | 1:51:33 | 9:16:25
368
+ [05:13:44] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000144
369
+ 1680 | 0 | 0.603608 | 0.785142 | 9.38e-06 | 1:52:10 | 9:15:33
370
+ [05:14:22] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
371
+ 1690 | 0 | 0.886907 | 0.788214 | 9.38e-06 | 1:52:48 | 9:14:40
372
+ [05:14:59] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000141
373
+ 1700 | 0 | 0.701825 | 0.789679 | 9.37e-06 | 1:53:25 | 9:13:48
374
+ [05:15:37] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000162
375
+ 1710 | 0 | 1.117721 | 0.795227 | 9.36e-06 | 1:54:03 | 9:12:55
376
+ [05:16:14] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000118
377
+ 1720 | 0 | 1.123572 | 0.800935 | 9.35e-06 | 1:54:41 | 9:12:05
378
+ [05:16:52] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000094
379
+ 1730 | 0 | 1.145399 | 0.803503 | 9.35e-06 | 1:55:18 | 9:11:13
380
+ [05:17:30] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000064
381
+ 1740 | 0 | 0.504704 | 0.802145 | 9.34e-06 | 1:55:56 | 9:10:22
382
+ [05:18:07] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000180
383
+ 1750 | 0 | 1.153842 | 0.802938 | 9.33e-06 | 1:56:33 | 9:09:31
384
+ [05:18:45] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000051
385
+ 1760 | 0 | 0.662771 | 0.804264 | 9.32e-06 | 1:57:11 | 9:08:40
386
+ [05:19:23] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000157
387
+ 1770 | 0 | 0.563092 | 0.804290 | 9.31e-06 | 1:57:49 | 9:07:50
388
+ [05:20:00] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000150
389
+ 1780 | 0 | 0.514744 | 0.798599 | 9.31e-06 | 1:58:26 | 9:06:59
390
+ [05:20:38] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000528
391
+ 1790 | 0 | 0.674077 | 0.796371 | 9.30e-06 | 1:59:04 | 9:06:09
392
+ [05:21:16] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000141
393
+ 1800 | 0 | 0.574359 | 0.793150 | 9.29e-06 | 1:59:42 | 9:05:19
394
+ [05:21:53] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000165
395
+ 1810 | 0 | 1.082740 | 0.797172 | 9.28e-06 | 2:00:19 | 9:04:29
396
+ [05:22:31] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000108
397
+ 1820 | 0 | 1.132584 | 0.796959 | 9.27e-06 | 2:00:57 | 9:03:38
398
+ [05:23:08] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000064
399
+ 1830 | 0 | 0.684599 | 0.797044 | 9.27e-06 | 2:01:34 | 9:02:47
400
+ [05:23:46] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000153
401
+ 1840 | 0 | 1.149022 | 0.803872 | 9.26e-06 | 2:02:11 | 9:01:53
402
+ [05:24:22] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000020
403
+ 1850 | 0 | 0.664640 | 0.798977 | 9.25e-06 | 2:02:49 | 9:01:03
404
+ [05:25:00] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000153
405
+ 1860 | 0 | 1.146904 | 0.800070 | 9.24e-06 | 2:03:27 | 9:00:15
406
+ [05:25:38] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000051
407
+ 1870 | 0 | 0.514053 | 0.798961 | 9.23e-06 | 2:04:04 | 8:59:27
408
+ [05:26:16] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000162
409
+ 1880 | 0 | 0.467661 | 0.795184 | 9.22e-06 | 2:04:42 | 8:58:38
410
+ [05:26:54] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000169
411
+ 1890 | 0 | 0.927803 | 0.794049 | 9.21e-06 | 2:05:20 | 8:57:49
412
+ [05:27:31] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000130
413
+ 1900 | 0 | 0.925098 | 0.798298 | 9.21e-06 | 2:05:56 | 8:56:56
414
+ [05:28:08] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000166
415
+ 1910 | 0 | 0.529765 | 0.799037 | 9.20e-06 | 2:06:32 | 8:56:00
416
+ [05:28:44] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000145
417
+ 1920 | 0 | 0.480941 | 0.798162 | 9.19e-06 | 2:07:08 | 8:55:04
418
+ [05:29:20] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000145
419
+ 1930 | 0 | 0.472238 | 0.791487 | 9.18e-06 | 2:07:46 | 8:54:14
420
+ [05:29:57] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000575
421
+ 1940 | 0 | 0.508702 | 0.785288 | 9.17e-06 | 2:08:23 | 8:53:26
422
+ [05:30:35] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000130
423
+ 1950 | 0 | 0.921684 | 0.788071 | 9.16e-06 | 2:09:00 | 8:52:33
424
+ [05:31:11] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000170
425
+ 1960 | 0 | 1.016356 | 0.791824 | 9.15e-06 | 2:09:37 | 8:51:44
426
+ [05:31:49] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000075
427
+ 1970 | 0 | 0.629489 | 0.786795 | 9.15e-06 | 2:10:15 | 8:50:56
428
+ [05:32:26] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000145
429
+ 1980 | 0 | 1.130274 | 0.792451 | 9.14e-06 | 2:10:53 | 8:50:08
430
+ [05:33:04] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000048
431
+ 1990 | 0 | 0.506427 | 0.786023 | 9.13e-06 | 2:11:30 | 8:49:21
432
+ [05:33:42] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000154
433
+ 2000 | 0 | 0.648079 | 0.781131 | 9.12e-06 | 2:12:08 | 8:48:34
434
+ [05:34:20] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000118
Magick/dlt/run_002.80c4d190-123e-452c-a6b0-0e8b7c55252f/validation/collage_step_0001000.png ADDED

Git LFS Details

  • SHA256: 429aa791a9edc779ef38f86f5e55417eb44794c623ce586aa859016d61fb1866
  • Pointer size: 133 Bytes
  • Size of remote file: 17.3 MB
Magick/dlt/run_003.21b84303-af7b-45c5-80da-56d08da757f8/checkpoint-2000/optimizer_0/__2_0.distcp ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c12e61a0303aadd96b2a259ab9b540b0dd9c87d4342fc488780de3a7e1a850ec
3
+ size 10335275880
Magick/dlt/run_003.21b84303-af7b-45c5-80da-56d08da757f8/checkpoint-2000/pytorch_model_fsdp_0/__2_0.distcp ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d4b9ef19f9bee2ec0e5f9ba05fb11739a7473ef022c98cba130abb462af17941
3
+ size 5167581436
Magick/dlt/run_003.21b84303-af7b-45c5-80da-56d08da757f8/checkpoint-2000/random_states_2.pkl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:95d691181dc63ad62f44607ca44e451fc2bd06ece6b32a1856f42a4da63073c8
3
+ size 14864
Magick/dlt/run_006/training.log ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Training started: 2026-08-26 16:46:26
2
+ ==========================================================================================
3
+ Step | Epoch | Loss | Avg100 | LR | Elapsed | ETA
4
+ ------------------------------------------------------------------------------------------
5
+ [16:46:26] Run directory: RESULTS/Magick/run_006
6
+ [16:46:26] Base model: black-forest-labs/FLUX.2-klein-base-4B
7
+ [16:46:26] cuDNN SDPA disabled -- using Flash/math backend for bf16 compatibility
8
+ [16:46:26] Loading pretrained FLUX.2-klein components ...
9
+ [16:46:32] transformer.config.guidance_embeds = False (guidance ignored (base/unguided model))
10
+ [16:46:39] Transformer params: 3,875,544,576 total, 3,875,544,576 trainable
11
+ [16:46:39] VAE scale factor: 8 (x2 for patchify = 16 pixels per packed token)
12
+ [16:46:39] TD ENABLED: cloning frozen EMA target transformer (ema_tau=0.999, update_freq=1 steps, k_idx=1.0, one_idx=0.3333333333333333, lambda=1.0, weighted=False, sigma_range=[0.002,1.0], rho=7.0, steps=18, sigma_data=1.0)
13
+ [16:46:44] Building dataset ...
14
+ [16:46:44] Step-based training: 10000 steps (~1 epochs over dataset)
15
+ [16:46:46]
16
+ [16:46:46] ================================================================
17
+ [16:46:46] Output dir = RESULTS/Magick/run_006
18
+ [16:46:46] Num examples = 74632
19
+ [16:46:46] Num epochs = 1
20
+ [16:46:46] Batch/device = 1
21
+ [16:46:46] Grad accum = 1
22
+ [16:46:46] Effective BS = 3
23
+ [16:46:46] Total optim steps = 10000
24
+ [16:46:46] LR = 1e-05
25
+ [16:46:46] Precision = bf16
26
+ [16:46:46] Text encoder layers= (9, 18, 27)
27
+ [16:46:46] Log every = 10 steps
28
+ [16:46:46] Validate every = 1000 steps
29
+ [16:46:46] Checkpoint every = 2000 steps
30
+ [16:46:46] TD = True
31
+ [16:46:46] ================================================================
32
+ [16:46:46]
Magick/dlt/run_007/training.log ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Training started: 2026-08-26 18:48:12
2
+ ==========================================================================================
3
+ Step | Epoch | Loss | Avg100 | LR | Elapsed | ETA
4
+ ------------------------------------------------------------------------------------------
5
+ [18:48:12] Run directory: RESULTS/Magick/run_007
6
+ [18:48:12] Base model: black-forest-labs/FLUX.2-klein-base-4B
7
+ [18:48:12] cuDNN SDPA disabled -- using Flash/math backend for bf16 compatibility
8
+ [18:48:12] Loading pretrained FLUX.2-klein components ...
9
+ [18:48:17] transformer.config.guidance_embeds = False (guidance ignored (base/unguided model))
10
+ [18:48:24] Transformer params: 3,875,544,576 total, 3,875,544,576 trainable
11
+ [18:48:24] VAE scale factor: 8 (x2 for patchify = 16 pixels per packed token)
12
+ [18:48:24] TD ENABLED: cloning frozen EMA target transformer (ema_tau=0.999, update_freq=1 steps, k_idx=1.0, one_idx=0.3333333333333333, lambda=1.0, weighted=False, sigma_range=[0.002,1.0], rho=7.0, steps=18, sigma_data=1.0)
13
+ [18:48:31] Building dataset ...
14
+ [18:48:31] Step-based training: 10000 steps (~1 epochs over dataset)
15
+ [18:48:42]
16
+ [18:48:42] ================================================================
17
+ [18:48:42] Output dir = RESULTS/Magick/run_007
18
+ [18:48:42] Num examples = 74632
19
+ [18:48:42] Num epochs = 1
20
+ [18:48:42] Batch/device = 1
21
+ [18:48:42] Grad accum = 1
22
+ [18:48:42] Effective BS = 3
23
+ [18:48:42] Total optim steps = 10000
24
+ [18:48:42] LR = 1e-05
25
+ [18:48:42] Precision = bf16
26
+ [18:48:42] Text encoder layers= (9, 18, 27)
27
+ [18:48:42] Log every = 10 steps
28
+ [18:48:42] Validate every = 1000 steps
29
+ [18:48:42] Checkpoint every = 2000 steps
30
+ [18:48:42] TD = True
31
+ [18:48:42] ================================================================
32
+ [18:48:42]
Magick/dlt/run_010/training.log ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Training started: 2026-08-26 18:52:06
2
+ ==========================================================================================
3
+ Step | Epoch | Loss | Avg100 | LR | Elapsed | ETA
4
+ ------------------------------------------------------------------------------------------
5
+ [18:52:06] Run directory: RESULTS/Magick/run_010
6
+ [18:52:06] Base model: black-forest-labs/FLUX.2-klein-base-4B
7
+ [18:52:06] cuDNN SDPA disabled -- using Flash/math backend for bf16 compatibility
8
+ [18:52:06] Loading pretrained FLUX.2-klein components ...
9
+ [18:52:11] transformer.config.guidance_embeds = False (guidance ignored (base/unguided model))
10
+ [18:52:18] Transformer params: 3,875,544,576 total, 3,875,544,576 trainable
11
+ [18:52:18] VAE scale factor: 8 (x2 for patchify = 16 pixels per packed token)
12
+ [18:52:18] TD ENABLED: cloning frozen EMA target transformer (ema_tau=0.999, update_freq=1 steps, k_idx=1.0, one_idx=0.3333333333333333, lambda=1.0, weighted=False, sigma_range=[0.002,1.0], rho=7.0, steps=18, sigma_data=1.0)
13
+ [18:52:24] Building dataset ...
14
+ [18:52:24] Step-based training: 10000 steps (~1 epochs over dataset)
15
+ [18:52:35]
16
+ [18:52:35] ================================================================
17
+ [18:52:35] Output dir = RESULTS/Magick/run_010
18
+ [18:52:35] Num examples = 74632
19
+ [18:52:35] Num epochs = 1
20
+ [18:52:35] Batch/device = 1
21
+ [18:52:35] Grad accum = 1
22
+ [18:52:35] Effective BS = 3
23
+ [18:52:35] Total optim steps = 10000
24
+ [18:52:35] LR = 1e-05
25
+ [18:52:35] Precision = bf16
26
+ [18:52:35] Text encoder layers= (9, 18, 27)
27
+ [18:52:35] Log every = 10 steps
28
+ [18:52:35] Validate every = 1000 steps
29
+ [18:52:35] Checkpoint every = 2000 steps
30
+ [18:52:35] TD = True
31
+ [18:52:35] ================================================================
32
+ [18:52:35]
Magick/dlt/run_016/training.log ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Training started: 2026-08-26 19:03:11
2
+ ==========================================================================================
3
+ Step | Epoch | Loss | Avg100 | LR | Elapsed | ETA
4
+ ------------------------------------------------------------------------------------------
5
+ [19:03:11] Run directory: RESULTS/Magick/run_016
6
+ [19:03:11] Base model: black-forest-labs/FLUX.2-klein-base-4B
7
+ [19:03:11] cuDNN SDPA disabled -- using Flash/math backend for bf16 compatibility
8
+ [19:03:11] Loading pretrained FLUX.2-klein components ...
9
+ [19:03:17] transformer.config.guidance_embeds = False (guidance ignored (base/unguided model))
10
+ [19:03:23] Transformer params: 3,875,544,576 total, 3,875,544,576 trainable
11
+ [19:03:23] VAE scale factor: 8 (x2 for patchify = 16 pixels per packed token)
12
+ [19:03:23] Building dataset ...
13
+ [19:03:23] Step-based training: 10000 steps (~1 epochs over dataset)
14
+ [19:03:32] TD ENABLED: building FSDP-sharded EMA target transformer (ema_tau=0.999, update_freq=1 steps, k_idx=1.0, one_idx=0.3333333333333333, lambda=1.0, weighted=False, sigma_range=[0.002,1.0], rho=7.0, steps=18, sigma_data=1.0)
Magick/dlt/run_019/training.log ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Training started: 2026-08-26 19:04:50
2
+ ==========================================================================================
3
+ Step | Epoch | Loss | Avg100 | LR | Elapsed | ETA
4
+ ------------------------------------------------------------------------------------------
5
+ [19:04:50] Run directory: RESULTS/Magick/run_019
6
+ [19:04:50] Base model: black-forest-labs/FLUX.2-klein-base-4B
7
+ [19:04:50] cuDNN SDPA disabled -- using Flash/math backend for bf16 compatibility
8
+ [19:04:50] Loading pretrained FLUX.2-klein components ...
9
+ [19:04:55] transformer.config.guidance_embeds = False (guidance ignored (base/unguided model))
10
+ [19:05:02] Transformer params: 3,875,544,576 total, 3,875,544,576 trainable
11
+ [19:05:02] VAE scale factor: 8 (x2 for patchify = 16 pixels per packed token)
12
+ [19:05:02] Building dataset ...
13
+ [19:05:02] Step-based training: 10000 steps (~1 epochs over dataset)
14
+ [19:05:21] TD ENABLED: building FSDP-sharded EMA target transformer (ema_tau=0.999, update_freq=1 steps, k_idx=1.0, one_idx=0.3333333333333333, lambda=1.0, weighted=False, sigma_range=[0.002,1.0], rho=7.0, steps=18, sigma_data=1.0)
Magick/dlt/run_023/training.log ADDED
@@ -0,0 +1,232 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Training started: 2026-08-26 19:06:26
2
+ ==========================================================================================
3
+ Step | Epoch | Loss | Avg100 | LR | Elapsed | ETA
4
+ ------------------------------------------------------------------------------------------
5
+ [19:06:26] Run directory: RESULTS/Magick/run_023
6
+ [19:06:26] Base model: black-forest-labs/FLUX.2-klein-base-4B
7
+ [19:06:26] cuDNN SDPA disabled -- using Flash/math backend for bf16 compatibility
8
+ [19:06:26] Loading pretrained FLUX.2-klein components ...
9
+ [19:06:30] transformer.config.guidance_embeds = False (guidance ignored (base/unguided model))
10
+ [19:06:37] Transformer params: 3,875,544,576 total, 3,875,544,576 trainable
11
+ [19:06:37] VAE scale factor: 8 (x2 for patchify = 16 pixels per packed token)
12
+ [19:06:37] Building dataset ...
13
+ [19:06:37] Step-based training: 10000 steps (~1 epochs over dataset)
14
+ [19:06:48] TD ENABLED: building FSDP-sharded EMA target transformer (ema_tau=0.999, update_freq=1 steps, k_idx=1.0, one_idx=0.3333333333333333, lambda=1.0, weighted=False, sigma_range=[0.002,1.0], rho=7.0, steps=18, sigma_data=1.0)
15
+ [19:06:50]
16
+ [19:06:50] ================================================================
17
+ [19:06:50] Output dir = RESULTS/Magick/run_023
18
+ [19:06:50] Num examples = 74632
19
+ [19:06:50] Num epochs = 1
20
+ [19:06:50] Batch/device = 1
21
+ [19:06:50] Grad accum = 1
22
+ [19:06:50] Effective BS = 3
23
+ [19:06:50] Total optim steps = 10000
24
+ [19:06:50] LR = 1e-05
25
+ [19:06:50] Precision = bf16
26
+ [19:06:50] Text encoder layers= (9, 18, 27)
27
+ [19:06:50] Log every = 10 steps
28
+ [19:06:50] Validate every = 1000 steps
29
+ [19:06:50] Checkpoint every = 2000 steps
30
+ [19:06:50] TD = True
31
+ [19:06:50] ================================================================
32
+ [19:06:50]
33
+ 10 | 0 | 1.328787 | 1.328787 | 1.00e-06 | 0:01:02 | 17:17:09
34
+ [19:07:28] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000078
35
+ 20 | 0 | 1.239188 | 1.283987 | 2.00e-06 | 0:01:39 | 13:47:51
36
+ [19:08:05] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000088
37
+ 30 | 0 | 1.110872 | 1.226282 | 3.00e-06 | 0:02:16 | 12:34:28
38
+ [19:08:42] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000053
39
+ 40 | 0 | 0.719512 | 1.099590 | 4.00e-06 | 0:02:53 | 12:00:16
40
+ [19:09:19] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000192
41
+ 50 | 0 | 0.529514 | 0.985575 | 5.00e-06 | 0:03:31 | 11:40:31
42
+ [19:09:57] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000254
43
+ 60 | 0 | 0.802840 | 0.955119 | 6.00e-06 | 0:04:08 | 11:25:33
44
+ [19:10:34] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000143
45
+ 70 | 0 | 0.543224 | 0.896277 | 7.00e-06 | 0:04:45 | 11:15:21
46
+ [19:11:11] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000189
47
+ 80 | 0 | 0.522107 | 0.849506 | 8.00e-06 | 0:05:23 | 11:09:12
48
+ [19:11:50] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000158
49
+ 90 | 0 | 0.587995 | 0.820449 | 9.00e-06 | 0:06:01 | 11:03:04
50
+ [19:12:27] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000164
51
+ 100 | 0 | 1.040190 | 0.842423 | 1.00e-05 | 0:06:38 | 10:58:02
52
+ [19:13:05] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000113
53
+ 110 | 0 | 0.861219 | 0.844132 | 1.00e-05 | 0:07:15 | 10:52:55
54
+ [19:13:42] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000158
55
+ 120 | 0 | 0.564767 | 0.820851 | 1.00e-05 | 0:07:51 | 10:47:33
56
+ [19:14:18] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000145
57
+ 130 | 0 | 0.472362 | 0.794044 | 1.00e-05 | 0:08:28 | 10:43:21
58
+ [19:14:54] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000397
59
+ 140 | 0 | 1.032136 | 0.811051 | 1.00e-05 | 0:09:05 | 10:40:48
60
+ [19:15:32] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000107
61
+ 150 | 0 | 1.079079 | 0.828919 | 1.00e-05 | 0:09:43 | 10:38:51
62
+ [19:16:10] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000094
63
+ 160 | 0 | 1.141186 | 0.848436 | 1.00e-05 | 0:10:21 | 10:36:44
64
+ [19:16:47] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000010
65
+ 170 | 0 | 0.522685 | 0.829274 | 1.00e-05 | 0:10:58 | 10:34:25
66
+ [19:17:24] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000817
67
+ 180 | 0 | 0.731419 | 0.823838 | 1.00e-05 | 0:11:36 | 10:33:06
68
+ [19:18:02] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000126
69
+ 190 | 0 | 0.473379 | 0.805393 | 1.00e-05 | 0:12:14 | 10:31:52
70
+ [19:18:40] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000164
71
+ 200 | 0 | 0.536389 | 0.791942 | 1.00e-05 | 0:12:51 | 10:30:19
72
+ [19:19:18] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000122
73
+ 210 | 0 | 0.677471 | 0.786491 | 1.00e-05 | 0:13:29 | 10:28:39
74
+ [19:19:55] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000157
75
+ 220 | 0 | 0.482186 | 0.772659 | 1.00e-05 | 0:14:07 | 10:27:48
76
+ [19:20:33] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000648
77
+ 230 | 0 | 0.513991 | 0.761413 | 1.00e-05 | 0:14:45 | 10:26:55
78
+ [19:21:11] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000195
79
+ 240 | 0 | 0.951258 | 0.769323 | 1.00e-05 | 0:15:23 | 10:25:42
80
+ [19:21:49] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000156
81
+ 250 | 0 | 0.861562 | 0.773013 | 9.99e-06 | 0:16:00 | 10:24:33
82
+ [19:22:27] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000119
83
+ 260 | 0 | 0.623910 | 0.767278 | 9.99e-06 | 0:16:37 | 10:23:05
84
+ [19:23:04] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000133
85
+ 270 | 0 | 0.655454 | 0.763136 | 9.99e-06 | 0:17:14 | 10:21:19
86
+ [19:23:40] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000140
87
+ 280 | 0 | 0.589771 | 0.756945 | 9.99e-06 | 0:17:50 | 10:19:34
88
+ [19:24:17] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000177
89
+ 290 | 0 | 0.881296 | 0.761233 | 9.99e-06 | 0:18:27 | 10:18:03
90
+ [19:24:53] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000146
91
+ 300 | 0 | 0.542805 | 0.753952 | 9.99e-06 | 0:19:05 | 10:17:12
92
+ [19:25:31] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000101
93
+ 310 | 0 | 0.786953 | 0.755016 | 9.99e-06 | 0:19:43 | 10:16:19
94
+ [19:26:09] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
95
+ 320 | 0 | 0.472776 | 0.746196 | 9.99e-06 | 0:20:20 | 10:15:23
96
+ [19:26:46] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000168
97
+ 330 | 0 | 0.768439 | 0.746870 | 9.99e-06 | 0:20:58 | 10:14:24
98
+ [19:27:24] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000192
99
+ 340 | 0 | 0.489814 | 0.739310 | 9.99e-06 | 0:21:34 | 10:13:08
100
+ [19:28:01] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000272
101
+ 350 | 0 | 0.832944 | 0.741985 | 9.98e-06 | 0:22:11 | 10:11:54
102
+ [19:28:37] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000151
103
+ 360 | 0 | 0.626334 | 0.738773 | 9.98e-06 | 0:22:49 | 10:11:06
104
+ [19:29:15] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000131
105
+ 370 | 0 | 0.936740 | 0.744123 | 9.98e-06 | 0:23:26 | 10:10:19
106
+ [19:29:53] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000170
107
+ 380 | 0 | 0.767685 | 0.744743 | 9.98e-06 | 0:24:04 | 10:09:20
108
+ [19:30:30] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000163
109
+ 390 | 0 | 0.703075 | 0.743675 | 9.98e-06 | 0:24:41 | 10:08:30
110
+ [19:31:08] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000131
111
+ 400 | 0 | 1.073708 | 0.751926 | 9.98e-06 | 0:25:19 | 10:07:50
112
+ [19:31:45] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000055
113
+ 410 | 0 | 0.597611 | 0.748162 | 9.98e-06 | 0:25:55 | 10:06:32
114
+ [19:32:22] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000126
115
+ 420 | 0 | 0.941781 | 0.752772 | 9.97e-06 | 0:26:33 | 10:05:52
116
+ [19:33:00] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000105
117
+ 430 | 0 | 0.475057 | 0.746313 | 9.97e-06 | 0:27:11 | 10:05:16
118
+ [19:33:38] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000175
119
+ 440 | 0 | 0.466102 | 0.739945 | 9.97e-06 | 0:27:49 | 10:04:38
120
+ [19:34:16] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000294
121
+ 450 | 0 | 0.544322 | 0.735598 | 9.97e-06 | 0:28:27 | 10:03:49
122
+ [19:34:53] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000153
123
+ 460 | 0 | 0.512960 | 0.730758 | 9.97e-06 | 0:29:04 | 10:03:01
124
+ [19:35:30] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000241
125
+ 470 | 0 | 0.579720 | 0.727544 | 9.97e-06 | 0:29:42 | 10:02:23
126
+ [19:36:08] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000160
127
+ 480 | 0 | 0.797737 | 0.729007 | 9.96e-06 | 0:30:20 | 10:01:36
128
+ [19:36:46] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000162
129
+ 490 | 0 | 0.828732 | 0.731042 | 9.96e-06 | 0:30:57 | 10:00:50
130
+ [19:37:23] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000159
131
+ 500 | 0 | 0.483678 | 0.726094 | 9.96e-06 | 0:31:34 | 10:00:02
132
+ [19:38:01] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000552
133
+ 510 | 0 | 1.140955 | 0.734229 | 9.96e-06 | 0:32:13 | 9:59:33
134
+ [19:38:39] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000073
135
+ 520 | 0 | 0.644179 | 0.732497 | 9.96e-06 | 0:32:51 | 9:58:53
136
+ [19:39:17] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000118
137
+ 530 | 0 | 0.496343 | 0.728041 | 9.95e-06 | 0:33:28 | 9:58:02
138
+ [19:39:54] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000131
139
+ 540 | 0 | 0.584294 | 0.725380 | 9.95e-06 | 0:34:04 | 9:57:05
140
+ [19:40:31] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000163
141
+ 550 | 0 | 0.989008 | 0.730173 | 9.95e-06 | 0:34:41 | 9:56:10
142
+ [19:41:08] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000108
143
+ 560 | 0 | 0.661321 | 0.728943 | 9.95e-06 | 0:35:19 | 9:55:31
144
+ [19:41:45] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000158
145
+ 570 | 0 | 0.485046 | 0.724664 | 9.94e-06 | 0:35:57 | 9:54:46
146
+ [19:42:23] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000560
147
+ 580 | 0 | 0.889146 | 0.727500 | 9.94e-06 | 0:36:35 | 9:54:11
148
+ [19:43:01] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000145
149
+ 590 | 0 | 1.081604 | 0.733502 | 9.94e-06 | 0:37:11 | 9:53:08
150
+ [19:43:37] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000100
151
+ 600 | 0 | 0.702014 | 0.732977 | 9.94e-06 | 0:37:48 | 9:52:24
152
+ [19:44:15] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000142
153
+ 610 | 0 | 0.780182 | 0.733751 | 9.93e-06 | 0:38:26 | 9:51:45
154
+ [19:44:52] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000199
155
+ 620 | 0 | 0.511294 | 0.730163 | 9.93e-06 | 0:39:04 | 9:51:14
156
+ [19:45:31] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000193
157
+ 630 | 0 | 0.631645 | 0.728599 | 9.93e-06 | 0:39:42 | 9:50:37
158
+ [19:46:08] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000138
159
+ 640 | 0 | 1.049505 | 0.733613 | 9.93e-06 | 0:40:19 | 9:49:52
160
+ [19:46:46] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000113
161
+ 650 | 0 | 0.852939 | 0.735449 | 9.92e-06 | 0:40:57 | 9:49:14
162
+ [19:47:24] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
163
+ 660 | 0 | 1.067133 | 0.740475 | 9.92e-06 | 0:41:35 | 9:48:37
164
+ [19:48:01] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000098
165
+ 670 | 0 | 0.948275 | 0.743576 | 9.92e-06 | 0:42:13 | 9:47:59
166
+ [19:48:39] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000151
167
+ 680 | 0 | 0.915175 | 0.746100 | 9.92e-06 | 0:42:50 | 9:47:16
168
+ [19:49:17] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000155
169
+ 690 | 0 | 0.579891 | 0.743691 | 9.91e-06 | 0:43:29 | 9:46:43
170
+ [19:49:55] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000125
171
+ 700 | 0 | 0.555466 | 0.741002 | 9.91e-06 | 0:44:06 | 9:46:03
172
+ [19:50:33] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000139
173
+ 710 | 0 | 0.561737 | 0.738477 | 9.91e-06 | 0:44:44 | 9:45:22
174
+ [19:51:10] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000142
175
+ 720 | 0 | 0.554218 | 0.735918 | 9.90e-06 | 0:45:21 | 9:44:33
176
+ [19:51:47] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000136
177
+ 730 | 0 | 0.889455 | 0.738021 | 9.90e-06 | 0:45:58 | 9:43:54
178
+ [19:52:25] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000175
179
+ 740 | 0 | 0.641531 | 0.736717 | 9.90e-06 | 0:46:36 | 9:43:12
180
+ [19:53:02] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000159
181
+ 750 | 0 | 1.075962 | 0.741241 | 9.89e-06 | 0:47:12 | 9:42:18
182
+ [19:53:39] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000109
183
+ 760 | 0 | 0.530438 | 0.738467 | 9.89e-06 | 0:47:50 | 9:41:39
184
+ [19:54:16] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000150
185
+ 770 | 0 | 0.560997 | 0.736162 | 9.89e-06 | 0:48:28 | 9:41:00
186
+ [19:54:54] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000138
187
+ 780 | 0 | 1.083259 | 0.740612 | 9.88e-06 | 0:49:06 | 9:40:23
188
+ [19:55:32] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000110
189
+ 790 | 0 | 0.897013 | 0.742592 | 9.88e-06 | 0:49:44 | 9:39:51
190
+ [19:56:10] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000172
191
+ 800 | 0 | 0.896392 | 0.744514 | 9.88e-06 | 0:50:22 | 9:39:14
192
+ [19:56:48] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000148
193
+ 810 | 0 | 0.680808 | 0.743728 | 9.87e-06 | 0:50:59 | 9:38:35
194
+ [19:57:26] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000149
195
+ 820 | 0 | 1.154063 | 0.748732 | 9.87e-06 | 0:51:37 | 9:37:56
196
+ [19:58:03] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000084
197
+ 830 | 0 | 0.676376 | 0.747860 | 9.87e-06 | 0:52:15 | 9:37:21
198
+ [19:58:41] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000162
199
+ 840 | 0 | 0.468061 | 0.744529 | 9.86e-06 | 0:52:53 | 9:36:41
200
+ [19:59:19] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000590
201
+ 850 | 0 | 1.155290 | 0.749362 | 9.86e-06 | 0:53:30 | 9:36:01
202
+ [19:59:56] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000039
203
+ 860 | 0 | 1.037725 | 0.752715 | 9.86e-06 | 0:54:08 | 9:35:24
204
+ [20:00:34] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000148
205
+ 870 | 0 | 0.624939 | 0.751246 | 9.85e-06 | 0:54:46 | 9:34:47
206
+ [20:01:12] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000134
207
+ 880 | 0 | 0.845804 | 0.752321 | 9.85e-06 | 0:55:23 | 9:34:07
208
+ [20:01:50] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000140
209
+ 890 | 0 | 1.041540 | 0.755570 | 9.84e-06 | 0:56:01 | 9:33:27
210
+ [20:02:27] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000117
211
+ 900 | 0 | 0.499844 | 0.752729 | 9.84e-06 | 0:56:38 | 9:32:44
212
+ [20:03:05] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000156
213
+ 910 | 0 | 0.455207 | 0.749459 | 9.84e-06 | 0:57:14 | 9:31:52
214
+ [20:03:41] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000320
215
+ 920 | 0 | 0.568698 | 0.747495 | 9.83e-06 | 0:57:51 | 9:31:00
216
+ [20:04:17] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000150
217
+ 930 | 0 | 1.139543 | 0.751710 | 9.83e-06 | 0:58:27 | 9:30:11
218
+ [20:04:54] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000075
219
+ 940 | 0 | 1.126432 | 0.755696 | 9.82e-06 | 0:59:06 | 9:29:38
220
+ [20:05:32] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000009
221
+ 950 | 0 | 0.669248 | 0.754786 | 9.82e-06 | 0:59:43 | 9:29:01
222
+ [20:06:10] TD diagnostics: frac_valid=1.000, td_loss_mean=0.001620
223
+ 960 | 0 | 0.640766 | 0.753599 | 9.81e-06 | 1:00:21 | 9:28:21
224
+ [20:06:47] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000129
225
+ 970 | 0 | 1.131469 | 0.757494 | 9.81e-06 | 1:00:59 | 9:27:46
226
+ [20:07:25] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000080
227
+ 980 | 0 | 0.564294 | 0.755523 | 9.81e-06 | 1:01:36 | 9:27:06
228
+ [20:08:03] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000124
229
+ 990 | 0 | 1.147596 | 0.759483 | 9.80e-06 | 1:02:14 | 9:26:27
230
+ [20:08:40] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000052
231
+ 1000 | 0 | 1.137585 | 0.763264 | 9.80e-06 | 1:02:52 | 9:25:50
232
+ [20:09:18] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000063
Magick/dlt/run_026/training.log ADDED
@@ -0,0 +1,233 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Training started: 2026-08-27 02:15:08
2
+ ==========================================================================================
3
+ Step | Epoch | Loss | Avg100 | LR | Elapsed | ETA
4
+ ------------------------------------------------------------------------------------------
5
+ [02:15:08] Run directory: RESULTS/Magick/run_026
6
+ [02:15:08] Base model: black-forest-labs/FLUX.2-klein-base-4B
7
+ [02:15:08] cuDNN SDPA disabled -- using Flash/math backend for bf16 compatibility
8
+ [02:15:08] Loading pretrained FLUX.2-klein components ...
9
+ [02:15:13] transformer.config.guidance_embeds = False (guidance ignored (base/unguided model))
10
+ [02:15:21] Transformer params: 3,875,544,576 total, 3,875,544,576 trainable
11
+ [02:15:21] VAE scale factor: 8 (x2 for patchify = 16 pixels per packed token)
12
+ [02:15:21] Building dataset ...
13
+ [02:15:21] Step-based training: 10000 steps (~1 epochs over dataset)
14
+ [02:15:29] TD ENABLED: building FSDP-sharded EMA target transformer (ema_tau=0.999, update_freq=1 steps, k_idx=1.0, one_idx=0.3333333333333333, lambda=1.0, weighted=False, sigma_range=[0.002,1.0], rho=7.0, steps=18, sigma_data=1.0)
15
+ [02:15:31]
16
+ [02:15:31] ================================================================
17
+ [02:15:31] Output dir = RESULTS/Magick/run_026
18
+ [02:15:31] Num examples = 74632
19
+ [02:15:31] Num epochs = 1
20
+ [02:15:31] Batch/device = 1
21
+ [02:15:31] Grad accum = 1
22
+ [02:15:31] Effective BS = 3
23
+ [02:15:31] Total optim steps = 10000
24
+ [02:15:31] LR = 1e-05
25
+ [02:15:31] Precision = bf16
26
+ [02:15:31] Text encoder layers= (9, 18, 27)
27
+ [02:15:31] Log every = 10 steps
28
+ [02:15:31] Validate every = 1000 steps
29
+ [02:15:31] Checkpoint every = 2000 steps
30
+ [02:15:31] TD = True
31
+ [02:15:31] ================================================================
32
+ [02:15:31]
33
+ 10 | 0 | 1.328054 | 1.328054 | 1.00e-06 | 0:01:01 | 17:11:55
34
+ [02:16:10] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000077
35
+ 20 | 0 | 1.239112 | 1.283583 | 2.00e-06 | 0:01:39 | 13:48:13
36
+ [02:16:47] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000087
37
+ 30 | 0 | 1.111912 | 1.226360 | 3.00e-06 | 0:02:17 | 12:40:53
38
+ [02:17:25] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000053
39
+ 40 | 0 | 0.718505 | 1.099396 | 4.00e-06 | 0:02:55 | 12:06:56
40
+ [02:18:03] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000186
41
+ 50 | 0 | 0.529058 | 0.985328 | 5.00e-06 | 0:03:33 | 11:46:41
42
+ [02:18:41] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000254
43
+ 60 | 0 | 0.802617 | 0.954877 | 6.00e-06 | 0:04:10 | 11:32:53
44
+ [02:19:19] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000142
45
+ 70 | 0 | 0.543135 | 0.896056 | 7.00e-06 | 0:04:48 | 11:22:22
46
+ [02:19:56] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000187
47
+ 80 | 0 | 0.521665 | 0.849258 | 8.00e-06 | 0:05:26 | 11:14:44
48
+ [02:20:34] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000158
49
+ 90 | 0 | 0.588598 | 0.820295 | 9.00e-06 | 0:06:04 | 11:08:11
50
+ [02:21:12] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000165
51
+ 100 | 0 | 1.040474 | 0.842313 | 1.00e-05 | 0:06:41 | 11:03:03
52
+ [02:21:50] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000111
53
+ 110 | 0 | 0.861257 | 0.844035 | 1.00e-05 | 0:07:19 | 10:58:51
54
+ [02:22:27] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000157
55
+ 120 | 0 | 0.564630 | 0.820752 | 1.00e-05 | 0:07:57 | 10:55:07
56
+ [02:23:05] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000146
57
+ 130 | 0 | 0.469026 | 0.793696 | 1.00e-05 | 0:08:33 | 10:50:11
58
+ [02:23:42] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000380
59
+ 140 | 0 | 1.032199 | 0.810732 | 1.00e-05 | 0:09:10 | 10:45:42
60
+ [02:24:18] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000107
61
+ 150 | 0 | 1.079298 | 0.828636 | 1.00e-05 | 0:09:48 | 10:43:48
62
+ [02:24:56] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000093
63
+ 160 | 0 | 1.138836 | 0.848024 | 1.00e-05 | 0:10:25 | 10:41:30
64
+ [02:25:34] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000010
65
+ 170 | 0 | 0.525772 | 0.829068 | 1.00e-05 | 0:11:03 | 10:39:34
66
+ [02:26:11] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000833
67
+ 180 | 0 | 0.731682 | 0.823657 | 1.00e-05 | 0:11:41 | 10:37:40
68
+ [02:26:49] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000128
69
+ 190 | 0 | 0.473540 | 0.805230 | 1.00e-05 | 0:12:19 | 10:36:04
70
+ [02:27:27] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000166
71
+ 200 | 0 | 0.536247 | 0.791781 | 1.00e-05 | 0:12:55 | 10:33:28
72
+ [02:28:03] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000120
73
+ 210 | 0 | 0.677148 | 0.786322 | 1.00e-05 | 0:13:32 | 10:31:37
74
+ [02:28:41] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000159
75
+ 220 | 0 | 0.483676 | 0.772566 | 1.00e-05 | 0:14:10 | 10:30:28
76
+ [02:29:19] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000665
77
+ 230 | 0 | 0.513898 | 0.761319 | 1.00e-05 | 0:14:48 | 10:29:07
78
+ [02:29:56] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000195
79
+ 240 | 0 | 0.951722 | 0.769253 | 1.00e-05 | 0:15:26 | 10:27:41
80
+ [02:30:34] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
81
+ 250 | 0 | 0.861186 | 0.772930 | 9.99e-06 | 0:16:03 | 10:26:26
82
+ [02:31:12] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000120
83
+ 260 | 0 | 0.624232 | 0.767211 | 9.99e-06 | 0:16:41 | 10:25:33
84
+ [02:31:50] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000135
85
+ 270 | 0 | 0.655068 | 0.763057 | 9.99e-06 | 0:17:18 | 10:23:46
86
+ [02:32:26] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000139
87
+ 280 | 0 | 0.589706 | 0.756866 | 9.99e-06 | 0:17:56 | 10:22:42
88
+ [02:33:04] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000176
89
+ 290 | 0 | 0.880863 | 0.761142 | 9.99e-06 | 0:18:33 | 10:21:31
90
+ [02:33:42] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000149
91
+ 300 | 0 | 0.542982 | 0.753870 | 9.99e-06 | 0:19:10 | 10:20:11
92
+ [02:34:19] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000102
93
+ 310 | 0 | 0.786500 | 0.754923 | 9.99e-06 | 0:19:48 | 10:19:01
94
+ [02:34:56] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
95
+ 320 | 0 | 0.472849 | 0.746108 | 9.99e-06 | 0:20:25 | 10:18:01
96
+ [02:35:34] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000168
97
+ 330 | 0 | 0.768561 | 0.746788 | 9.99e-06 | 0:21:03 | 10:16:50
98
+ [02:36:11] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000194
99
+ 340 | 0 | 0.488401 | 0.739189 | 9.99e-06 | 0:21:40 | 10:15:53
100
+ [02:36:48] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000270
101
+ 350 | 0 | 0.832232 | 0.741847 | 9.98e-06 | 0:22:18 | 10:14:53
102
+ [02:37:26] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000155
103
+ 360 | 0 | 0.625617 | 0.738618 | 9.98e-06 | 0:22:54 | 10:13:30
104
+ [02:38:02] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000127
105
+ 370 | 0 | 0.936859 | 0.743976 | 9.98e-06 | 0:23:31 | 10:12:12
106
+ [02:38:39] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000168
107
+ 380 | 0 | 0.767635 | 0.744599 | 9.98e-06 | 0:24:08 | 10:11:13
108
+ [02:39:16] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000163
109
+ 390 | 0 | 0.703094 | 0.743535 | 9.98e-06 | 0:24:46 | 10:10:25
110
+ [02:39:54] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000132
111
+ 400 | 0 | 1.073391 | 0.751781 | 9.98e-06 | 0:25:24 | 10:09:39
112
+ [02:40:32] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000054
113
+ 410 | 0 | 0.597818 | 0.748026 | 9.98e-06 | 0:26:01 | 10:08:51
114
+ [02:41:10] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000126
115
+ 420 | 0 | 0.941953 | 0.752643 | 9.97e-06 | 0:26:39 | 10:08:02
116
+ [02:41:47] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000105
117
+ 430 | 0 | 0.474975 | 0.746186 | 9.97e-06 | 0:27:17 | 10:07:14
118
+ [02:42:25] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000173
119
+ 440 | 0 | 0.465445 | 0.739805 | 9.97e-06 | 0:27:55 | 10:06:34
120
+ [02:43:03] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000293
121
+ 450 | 0 | 0.544741 | 0.735471 | 9.97e-06 | 0:28:33 | 10:05:54
122
+ [02:43:41] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
123
+ 460 | 0 | 0.513572 | 0.730647 | 9.97e-06 | 0:29:10 | 10:05:00
124
+ [02:44:18] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000246
125
+ 470 | 0 | 0.579491 | 0.727431 | 9.97e-06 | 0:29:48 | 10:04:20
126
+ [02:44:56] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000156
127
+ 480 | 0 | 0.798705 | 0.728915 | 9.96e-06 | 0:30:25 | 10:03:31
128
+ [02:45:34] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000163
129
+ 490 | 0 | 0.828976 | 0.730958 | 9.96e-06 | 0:31:03 | 10:02:43
130
+ [02:46:11] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000161
131
+ 500 | 0 | 0.482081 | 0.725980 | 9.96e-06 | 0:31:40 | 10:01:52
132
+ [02:46:48] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000545
133
+ 510 | 0 | 1.141370 | 0.734125 | 9.96e-06 | 0:32:18 | 10:01:12
134
+ [02:47:26] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000072
135
+ 520 | 0 | 0.644343 | 0.732398 | 9.96e-06 | 0:32:55 | 10:00:19
136
+ [02:48:04] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000117
137
+ 530 | 0 | 0.495395 | 0.727927 | 9.95e-06 | 0:33:31 | 9:59:06
138
+ [02:48:40] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000132
139
+ 540 | 0 | 0.584211 | 0.725265 | 9.95e-06 | 0:34:08 | 9:58:09
140
+ [02:49:16] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000162
141
+ 550 | 0 | 0.985051 | 0.729989 | 9.95e-06 | 0:34:45 | 9:57:12
142
+ [02:49:53] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000113
143
+ 560 | 0 | 0.660874 | 0.728754 | 9.95e-06 | 0:35:23 | 9:56:28
144
+ [02:50:31] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000152
145
+ 570 | 0 | 0.488672 | 0.724542 | 9.94e-06 | 0:36:00 | 9:55:47
146
+ [02:51:09] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000577
147
+ 580 | 0 | 0.889000 | 0.727378 | 9.94e-06 | 0:36:38 | 9:55:10
148
+ [02:51:47] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000146
149
+ 590 | 0 | 1.081410 | 0.733378 | 9.94e-06 | 0:37:16 | 9:54:29
150
+ [02:52:24] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000102
151
+ 600 | 0 | 0.701982 | 0.732855 | 9.94e-06 | 0:37:54 | 9:53:46
152
+ [02:53:02] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000144
153
+ 610 | 0 | 0.780209 | 0.733631 | 9.93e-06 | 0:38:30 | 9:52:48
154
+ [02:53:38] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000202
155
+ 620 | 0 | 0.511497 | 0.730049 | 9.93e-06 | 0:39:08 | 9:52:13
156
+ [02:54:16] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000192
157
+ 630 | 0 | 0.631430 | 0.728483 | 9.93e-06 | 0:39:45 | 9:51:13
158
+ [02:54:53] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000139
159
+ 640 | 0 | 1.049324 | 0.733496 | 9.93e-06 | 0:40:21 | 9:50:07
160
+ [02:55:29] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000113
161
+ 650 | 0 | 0.853095 | 0.735336 | 9.92e-06 | 0:40:57 | 9:49:11
162
+ [02:56:05] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000150
163
+ 660 | 0 | 1.067921 | 0.740376 | 9.92e-06 | 0:41:33 | 9:48:11
164
+ [02:56:42] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000097
165
+ 670 | 0 | 0.948980 | 0.743489 | 9.92e-06 | 0:42:10 | 9:47:16
166
+ [02:57:18] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000149
167
+ 680 | 0 | 0.914990 | 0.746011 | 9.92e-06 | 0:42:46 | 9:46:12
168
+ [02:57:54] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000155
169
+ 690 | 0 | 0.579871 | 0.743603 | 9.91e-06 | 0:43:22 | 9:45:19
170
+ [02:58:31] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000125
171
+ 700 | 0 | 0.555968 | 0.740923 | 9.91e-06 | 0:43:58 | 9:44:18
172
+ [02:59:07] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000141
173
+ 710 | 0 | 0.563465 | 0.738423 | 9.91e-06 | 0:44:35 | 9:43:32
174
+ [02:59:44] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000145
175
+ 720 | 0 | 0.555232 | 0.735879 | 9.90e-06 | 0:45:13 | 9:42:54
176
+ [03:00:21] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000140
177
+ 730 | 0 | 0.890111 | 0.737992 | 9.90e-06 | 0:45:50 | 9:42:06
178
+ [03:00:58] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000172
179
+ 740 | 0 | 0.642775 | 0.736705 | 9.90e-06 | 0:46:27 | 9:41:25
180
+ [03:01:36] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000163
181
+ 750 | 0 | 1.077138 | 0.741244 | 9.89e-06 | 0:47:05 | 9:40:47
182
+ [03:02:13] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000110
183
+ 760 | 0 | 0.530762 | 0.738475 | 9.89e-06 | 0:47:43 | 9:40:15
184
+ [03:02:51] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000150
185
+ 770 | 0 | 0.560388 | 0.736162 | 9.89e-06 | 0:48:21 | 9:39:37
186
+ [03:03:29] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000137
187
+ 780 | 0 | 1.082966 | 0.740608 | 9.88e-06 | 0:48:58 | 9:38:59
188
+ [03:04:07] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000113
189
+ 790 | 0 | 0.896653 | 0.742583 | 9.88e-06 | 0:49:36 | 9:38:18
190
+ [03:04:44] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000178
191
+ 800 | 0 | 0.895721 | 0.744498 | 9.88e-06 | 0:50:13 | 9:37:38
192
+ [03:05:22] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000147
193
+ 810 | 0 | 0.680525 | 0.743708 | 9.87e-06 | 0:50:51 | 9:36:59
194
+ [03:05:59] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000149
195
+ 820 | 0 | 1.154117 | 0.748713 | 9.87e-06 | 0:51:28 | 9:36:20
196
+ [03:06:37] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000084
197
+ 830 | 0 | 0.676429 | 0.747842 | 9.87e-06 | 0:52:06 | 9:35:44
198
+ [03:07:14] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000160
199
+ 840 | 0 | 0.473974 | 0.744582 | 9.86e-06 | 0:52:43 | 9:35:01
200
+ [03:07:52] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000595
201
+ 850 | 0 | 1.153181 | 0.749389 | 9.86e-06 | 0:53:21 | 9:34:20
202
+ [03:08:29] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000038
203
+ 860 | 0 | 1.036955 | 0.752732 | 9.86e-06 | 0:53:58 | 9:33:41
204
+ [03:09:07] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000147
205
+ 870 | 0 | 0.624921 | 0.751263 | 9.85e-06 | 0:54:36 | 9:33:05
206
+ [03:09:44] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000133
207
+ 880 | 0 | 0.845646 | 0.752336 | 9.85e-06 | 0:55:14 | 9:32:26
208
+ [03:10:22] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000143
209
+ 890 | 0 | 1.041445 | 0.755584 | 9.84e-06 | 0:55:51 | 9:31:44
210
+ [03:10:59] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000114
211
+ 900 | 0 | 0.499893 | 0.752743 | 9.84e-06 | 0:56:29 | 9:31:11
212
+ [03:11:37] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000156
213
+ 910 | 0 | 0.454727 | 0.749468 | 9.84e-06 | 0:57:07 | 9:30:33
214
+ [03:12:15] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000317
215
+ 920 | 0 | 0.568276 | 0.747499 | 9.83e-06 | 0:57:44 | 9:29:51
216
+ [03:12:52] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000145
217
+ 930 | 0 | 1.139695 | 0.751716 | 9.83e-06 | 0:58:21 | 9:29:05
218
+ [03:13:29] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000075
219
+ 940 | 0 | 1.128800 | 0.755728 | 9.82e-06 | 0:58:58 | 9:28:29
220
+ [03:14:07] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000010
221
+ 950 | 0 | 0.646154 | 0.754574 | 9.82e-06 | 0:59:36 | 9:27:51
222
+ [03:14:44] TD diagnostics: frac_valid=1.000, td_loss_mean=0.001489
223
+ 960 | 0 | 0.641101 | 0.753392 | 9.81e-06 | 1:00:14 | 9:27:14
224
+ [03:15:22] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000132
225
+ 970 | 0 | 1.131585 | 0.757291 | 9.81e-06 | 1:00:51 | 9:26:33
226
+ [03:15:59] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000081
227
+ 980 | 0 | 0.565031 | 0.755329 | 9.81e-06 | 1:01:27 | 9:25:40
228
+ [03:16:35] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000123
229
+ 990 | 0 | 1.146037 | 0.759276 | 9.80e-06 | 1:02:05 | 9:25:01
230
+ [03:17:13] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000051
231
+ 1000 | 0 | 1.137030 | 0.763053 | 9.80e-06 | 1:02:42 | 9:24:23
232
+ [03:17:50] TD diagnostics: frac_valid=1.000, td_loss_mean=0.000063
233
+ [03:18:20] Generating 16 validation images ...
Magick/flux 2/Baseline/checkpoint-10000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:017af2292c2715906e71ccba6c56e277007cef44024157e781262b1d9c5b158d
3
+ size 7751109744
Magick/flux 2/Baseline/checkpoint-10000/optimizer.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:670f1e89cca5fea441b7913b6ab24ca2d1547e4a4703dd981072c79d2c8827d3
3
+ size 7872415698
Magick/flux 2/Baseline/checkpoint-10000/random_states_0.pkl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f36248957dbfd046b7e131a7a8e944821ca50c384ca8366f1ef06f51f790f077
3
+ size 14864
Magick/flux 2/Baseline/checkpoint-10000/scheduler.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:292caa766a63ee276b2e48363769212865b7fc5884f22f4e3b18dce828a3fe59
3
+ size 1000
Magick/flux 2/Baseline/checkpoint-10000/transformer/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "Flux2Transformer2DModel",
3
+ "_diffusers_version": "0.37.1",
4
+ "_name_or_path": "black-forest-labs/FLUX.2-klein-base-4B",
5
+ "attention_head_dim": 128,
6
+ "axes_dims_rope": [
7
+ 32,
8
+ 32,
9
+ 32,
10
+ 32
11
+ ],
12
+ "eps": 1e-06,
13
+ "guidance_embeds": false,
14
+ "in_channels": 128,
15
+ "joint_attention_dim": 7680,
16
+ "mlp_ratio": 3.0,
17
+ "num_attention_heads": 24,
18
+ "num_layers": 5,
19
+ "num_single_layers": 20,
20
+ "out_channels": null,
21
+ "patch_size": 1,
22
+ "rope_theta": 2000,
23
+ "timestep_guidance_channels": 256
24
+ }
Magick/flux 2/Baseline/checkpoint-10000/transformer/diffusion_pytorch_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:017af2292c2715906e71ccba6c56e277007cef44024157e781262b1d9c5b158d
3
+ size 7751109744
Magick/flux 2/Baseline/checkpoint-8000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ee2599aed6f78acc1f10780b8f229e31435259fcc14435458bef5ceb04b31d0
3
+ size 7751109744
Magick/flux 2/Baseline/checkpoint-8000/optimizer.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a9350e50688cccadac4ac74f90576d66ad8cef0b344eb2cacac495ce7052c26
3
+ size 7872415698
Magick/flux 2/Baseline/checkpoint-8000/random_states_0.pkl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:401d41582081d2da9d41f3f9b8101746d371f51061f9b869954c0fa16764cb0d
3
+ size 14864
Magick/flux 2/Baseline/checkpoint-8000/scheduler.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2bc54396495395cef025a33eb1d9f1c722c2e99c89815036ab41bc84e1aff41f
3
+ size 1000
Magick/flux 2/Baseline/checkpoint-8000/transformer/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "Flux2Transformer2DModel",
3
+ "_diffusers_version": "0.37.1",
4
+ "_name_or_path": "black-forest-labs/FLUX.2-klein-base-4B",
5
+ "attention_head_dim": 128,
6
+ "axes_dims_rope": [
7
+ 32,
8
+ 32,
9
+ 32,
10
+ 32
11
+ ],
12
+ "eps": 1e-06,
13
+ "guidance_embeds": false,
14
+ "in_channels": 128,
15
+ "joint_attention_dim": 7680,
16
+ "mlp_ratio": 3.0,
17
+ "num_attention_heads": 24,
18
+ "num_layers": 5,
19
+ "num_single_layers": 20,
20
+ "out_channels": null,
21
+ "patch_size": 1,
22
+ "rope_theta": 2000,
23
+ "timestep_guidance_channels": 256
24
+ }
Magick/flux 2/Baseline/checkpoint-8000/transformer/diffusion_pytorch_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ee2599aed6f78acc1f10780b8f229e31435259fcc14435458bef5ceb04b31d0
3
+ size 7751109744
Magick/flux 2/Baseline/checkpoint-9000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4bc77b9f9d3f2f53dfb8b2da889871aa55b1b673a5924d7601329013bbad7e4e
3
+ size 7751109744
Magick/flux 2/Baseline/checkpoint-9000/optimizer.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7b4f2e1f50700271449bc881344abab64da3301c91575912f8b1af5aba1843b
3
+ size 7872415698
Magick/flux 2/Baseline/checkpoint-9000/random_states_0.pkl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59be7bfc1a44047e90a40d9477534c352c6fa4dedb5529146feebdfa268eaed4
3
+ size 14864
Magick/flux 2/Baseline/checkpoint-9000/scheduler.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a7d33e55551f05c1d81ec8ff4e98a1894e57b39d39f389462f6b5b02d4fee25b
3
+ size 1000
Magick/flux 2/Baseline/checkpoint-9000/transformer/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "Flux2Transformer2DModel",
3
+ "_diffusers_version": "0.37.1",
4
+ "_name_or_path": "black-forest-labs/FLUX.2-klein-base-4B",
5
+ "attention_head_dim": 128,
6
+ "axes_dims_rope": [
7
+ 32,
8
+ 32,
9
+ 32,
10
+ 32
11
+ ],
12
+ "eps": 1e-06,
13
+ "guidance_embeds": false,
14
+ "in_channels": 128,
15
+ "joint_attention_dim": 7680,
16
+ "mlp_ratio": 3.0,
17
+ "num_attention_heads": 24,
18
+ "num_layers": 5,
19
+ "num_single_layers": 20,
20
+ "out_channels": null,
21
+ "patch_size": 1,
22
+ "rope_theta": 2000,
23
+ "timestep_guidance_channels": 256
24
+ }
Magick/flux 2/Baseline/checkpoint-9000/transformer/diffusion_pytorch_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4bc77b9f9d3f2f53dfb8b2da889871aa55b1b673a5924d7601329013bbad7e4e
3
+ size 7751109744
Magick/flux 2/Baseline/model_index.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "Flux2KleinPipeline",
3
+ "_diffusers_version": "0.37.1",
4
+ "is_distilled": false,
5
+ "scheduler": [
6
+ "diffusers",
7
+ "FlowMatchEulerDiscreteScheduler"
8
+ ],
9
+ "text_encoder": [
10
+ "transformers",
11
+ "Qwen3ForCausalLM"
12
+ ],
13
+ "tokenizer": [
14
+ "transformers",
15
+ "Qwen2Tokenizer"
16
+ ],
17
+ "transformer": [
18
+ "diffusers",
19
+ "Flux2Transformer2DModel"
20
+ ],
21
+ "vae": [
22
+ "diffusers",
23
+ "AutoencoderKLFlux2"
24
+ ]
25
+ }
Magick/flux 2/Baseline/scheduler/scheduler_config.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "FlowMatchEulerDiscreteScheduler",
3
+ "_diffusers_version": "0.37.1",
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
+ }
Magick/flux 2/Baseline/text_encoder/config.json ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3ForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 151643,
8
+ "dtype": "bfloat16",
9
+ "eos_token_id": 151645,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2560,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 9728,
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": 40960,
54
+ "max_window_layers": 36,
55
+ "model_type": "qwen3",
56
+ "num_attention_heads": 32,
57
+ "num_hidden_layers": 36,
58
+ "num_key_value_heads": 8,
59
+ "pad_token_id": null,
60
+ "rms_norm_eps": 1e-06,
61
+ "rope_parameters": {
62
+ "rope_theta": 1000000,
63
+ "rope_type": "default"
64
+ },
65
+ "sliding_window": null,
66
+ "tie_word_embeddings": true,
67
+ "transformers_version": "5.9.0",
68
+ "use_cache": true,
69
+ "use_sliding_window": false,
70
+ "vocab_size": 151936
71
+ }
Magick/flux 2/Baseline/text_encoder/generation_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 151643,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "pad_token_id": 151643,
9
+ "temperature": 0.6,
10
+ "top_k": 20,
11
+ "top_p": 0.95,
12
+ "transformers_version": "5.9.0"
13
+ }
Magick/flux 2/Baseline/text_encoder/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca92a98161ee65bb01d2788fbca632e105460f72df4bf010e3212f5174ec7e6f
3
+ size 8044982080
Magick/flux 2/Baseline/tokenizer/chat_template.jinja ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {{- messages[0].content + '\n\n' }}
5
+ {%- endif %}
6
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
7
+ {%- for tool in tools %}
8
+ {{- "\n" }}
9
+ {{- tool | tojson }}
10
+ {%- endfor %}
11
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
12
+ {%- else %}
13
+ {%- if messages[0].role == 'system' %}
14
+ {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
15
+ {%- endif %}
16
+ {%- endif %}
17
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
18
+ {%- for message in messages[::-1] %}
19
+ {%- set index = (messages|length - 1) - loop.index0 %}
20
+ {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
21
+ {%- set ns.multi_step_tool = false %}
22
+ {%- set ns.last_query_index = index %}
23
+ {%- endif %}
24
+ {%- endfor %}
25
+ {%- for message in messages %}
26
+ {%- if message.content is string %}
27
+ {%- set content = message.content %}
28
+ {%- else %}
29
+ {%- set content = '' %}
30
+ {%- endif %}
31
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
32
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
33
+ {%- elif message.role == "assistant" %}
34
+ {%- set reasoning_content = '' %}
35
+ {%- if message.reasoning_content is string %}
36
+ {%- set reasoning_content = message.reasoning_content %}
37
+ {%- else %}
38
+ {%- if '</think>' in content %}
39
+ {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
40
+ {%- set content = content.split('</think>')[-1].lstrip('\n') %}
41
+ {%- endif %}
42
+ {%- endif %}
43
+ {%- if loop.index0 > ns.last_query_index %}
44
+ {%- if loop.last or (not loop.last and reasoning_content) %}
45
+ {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
46
+ {%- else %}
47
+ {{- '<|im_start|>' + message.role + '\n' + content }}
48
+ {%- endif %}
49
+ {%- else %}
50
+ {{- '<|im_start|>' + message.role + '\n' + content }}
51
+ {%- endif %}
52
+ {%- if message.tool_calls %}
53
+ {%- for tool_call in message.tool_calls %}
54
+ {%- if (loop.first and content) or (not loop.first) %}
55
+ {{- '\n' }}
56
+ {%- endif %}
57
+ {%- if tool_call.function %}
58
+ {%- set tool_call = tool_call.function %}
59
+ {%- endif %}
60
+ {{- '<tool_call>\n{"name": "' }}
61
+ {{- tool_call.name }}
62
+ {{- '", "arguments": ' }}
63
+ {%- if tool_call.arguments is string %}
64
+ {{- tool_call.arguments }}
65
+ {%- else %}
66
+ {{- tool_call.arguments | tojson }}
67
+ {%- endif %}
68
+ {{- '}\n</tool_call>' }}
69
+ {%- endfor %}
70
+ {%- endif %}
71
+ {{- '<|im_end|>\n' }}
72
+ {%- elif message.role == "tool" %}
73
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
74
+ {{- '<|im_start|>user' }}
75
+ {%- endif %}
76
+ {{- '\n<tool_response>\n' }}
77
+ {{- content }}
78
+ {{- '\n</tool_response>' }}
79
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
80
+ {{- '<|im_end|>\n' }}
81
+ {%- endif %}
82
+ {%- endif %}
83
+ {%- endfor %}
84
+ {%- if add_generation_prompt %}
85
+ {{- '<|im_start|>assistant\n' }}
86
+ {%- if enable_thinking is defined and enable_thinking is false %}
87
+ {{- '<think>\n\n</think>\n\n' }}
88
+ {%- endif %}
89
+ {%- endif %}
Magick/flux 2/Baseline/tokenizer/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ebc61ca942af8e084a5c98257c3c21042d76a2e8c8324fc242fd19976cb987a
3
+ size 11422928
Magick/flux 2/Baseline/tokenizer/tokenizer_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "is_local": false,
9
+ "local_files_only": false,
10
+ "model_max_length": 131072,
11
+ "pad_token": "<|endoftext|>",
12
+ "split_special_tokens": false,
13
+ "tokenizer_class": "Qwen2Tokenizer",
14
+ "unk_token": null
15
+ }