Upload folder using huggingface_hub
Browse files- README.md +9 -11
- text_encoder/0.safetensors +2 -2
- text_encoder/1.safetensors +2 -2
- text_encoder/model.safetensors.index.json +44 -44
- transformer/0.safetensors +2 -2
- transformer/1.safetensors +2 -2
- transformer/2.safetensors +2 -2
- transformer/3.safetensors +3 -0
- transformer/model.safetensors.index.json +605 -605
- vae/0.safetensors +1 -1
- vae/model.safetensors.index.json +1 -1
README.md
CHANGED
|
@@ -8,14 +8,12 @@ tags:
|
|
| 8 |
- mlx-gen
|
| 9 |
- mflux
|
| 10 |
- apple-silicon
|
| 11 |
-
-
|
| 12 |
-
- mixed-q4
|
| 13 |
-
- mixed-q4-q8
|
| 14 |
- ernie
|
| 15 |
- ernie-image
|
| 16 |
- ernie-image-turbo
|
| 17 |
---
|
| 18 |
-
# ernie-image-turbo-
|
| 19 |
|
| 20 |
This repository contains MLX-Gen saved weights for `baidu/ERNIE-Image-Turbo`. The checkpoint is designed for local Apple Silicon inference with [`mlx-gen`](https://github.com/lpalbou/mlx-gen).
|
| 21 |
|
|
@@ -31,15 +29,15 @@ This quantized derivative follows the Apache 2.0 license of the source model.
|
|
| 31 |
|
| 32 |
## Quantization
|
| 33 |
|
| 34 |
-
This is an MLX
|
| 35 |
|
| 36 |
-
-
|
| 37 |
-
- q8 for ERNIE
|
| 38 |
-
- q8 for ERNIE text projection, timestep embedding, AdaLN modulation, final norm, and final projection.
|
| 39 |
-
- q8 for Mistral3 text-encoder and Prompt Enhancer linears.
|
| 40 |
- q8 for quantizable ERNIE VAE attention modules.
|
| 41 |
- BF16 for norms, convolutions, and other non-quantizable parameters.
|
| 42 |
|
|
|
|
|
|
|
| 43 |
See the [MLX-Gen quantization docs](https://github.com/lpalbou/mlx-gen/blob/main/docs/quantization.md) for compatibility notes and measured ERNIE q4/q8 behavior.
|
| 44 |
|
| 45 |
Prepared ERNIE folders contain the ordinary text-to-image generation stack. ERNIE Prompt Enhancer files are not bundled in this checkpoint.
|
|
@@ -57,10 +55,10 @@ Use the `mlxgen` command and Python import path for new MLX-Gen projects.
|
|
| 57 |
```bash
|
| 58 |
python -m pip install -U mlx-gen
|
| 59 |
|
| 60 |
-
mlxgen download --model AbstractFramework/ernie-image-turbo-
|
| 61 |
|
| 62 |
mlxgen generate \
|
| 63 |
-
--model AbstractFramework/ernie-image-turbo-
|
| 64 |
--prompt "Your prompt here" \
|
| 65 |
--width 512 \
|
| 66 |
--height 512 \
|
|
|
|
| 8 |
- mlx-gen
|
| 9 |
- mflux
|
| 10 |
- apple-silicon
|
| 11 |
+
- 8-bit
|
|
|
|
|
|
|
| 12 |
- ernie
|
| 13 |
- ernie-image
|
| 14 |
- ernie-image-turbo
|
| 15 |
---
|
| 16 |
+
# ernie-image-turbo-8bit
|
| 17 |
|
| 18 |
This repository contains MLX-Gen saved weights for `baidu/ERNIE-Image-Turbo`. The checkpoint is designed for local Apple Silicon inference with [`mlx-gen`](https://github.com/lpalbou/mlx-gen).
|
| 19 |
|
|
|
|
| 29 |
|
| 30 |
## Quantization
|
| 31 |
|
| 32 |
+
This is an MLX q8 checkpoint for ERNIE Image Turbo. MLX-Gen uses 8-bit quantization for ERNIE modules where MLX supports quantization:
|
| 33 |
|
| 34 |
+
- q8 for quantizable ERNIE transformer modules.
|
| 35 |
+
- q8 for quantizable ERNIE text-encoder modules.
|
|
|
|
|
|
|
| 36 |
- q8 for quantizable ERNIE VAE attention modules.
|
| 37 |
- BF16 for norms, convolutions, and other non-quantizable parameters.
|
| 38 |
|
| 39 |
+
ERNIE q4 uses a model-specific mixed q4/q8 policy because fully q4 checkpoints can drift from BF16/q8 behavior.
|
| 40 |
+
|
| 41 |
See the [MLX-Gen quantization docs](https://github.com/lpalbou/mlx-gen/blob/main/docs/quantization.md) for compatibility notes and measured ERNIE q4/q8 behavior.
|
| 42 |
|
| 43 |
Prepared ERNIE folders contain the ordinary text-to-image generation stack. ERNIE Prompt Enhancer files are not bundled in this checkpoint.
|
|
|
|
| 55 |
```bash
|
| 56 |
python -m pip install -U mlx-gen
|
| 57 |
|
| 58 |
+
mlxgen download --model AbstractFramework/ernie-image-turbo-8bit
|
| 59 |
|
| 60 |
mlxgen generate \
|
| 61 |
+
--model AbstractFramework/ernie-image-turbo-8bit \
|
| 62 |
--prompt "Your prompt here" \
|
| 63 |
--width 512 \
|
| 64 |
--height 512 \
|
text_encoder/0.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e11ae4fb412f397d2496db26ddeb712d724650df3f9608a6889f0cc015651398
|
| 3 |
+
size 2129274425
|
text_encoder/1.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2af0b3bb912474ecee1c7d8858658736c578cd3bbe278cbd76ccecde9da1d14
|
| 3 |
+
size 1514262266
|
text_encoder/model.safetensors.index.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"metadata": {
|
| 3 |
-
"quantization_level": "
|
| 4 |
"mflux_version": "0.18.5"
|
| 5 |
},
|
| 6 |
"weight_map": {
|
|
@@ -326,49 +326,49 @@
|
|
| 326 |
"layers.13.mlp.up_proj.weight": "0.safetensors",
|
| 327 |
"layers.13.mlp.up_proj.scales": "0.safetensors",
|
| 328 |
"layers.13.mlp.up_proj.biases": "0.safetensors",
|
| 329 |
-
"layers.13.mlp.down_proj.weight": "
|
| 330 |
-
"layers.13.mlp.down_proj.scales": "
|
| 331 |
-
"layers.13.mlp.down_proj.biases": "
|
| 332 |
-
"layers.14.input_layernorm.weight": "
|
| 333 |
-
"layers.14.post_attention_layernorm.weight": "
|
| 334 |
-
"layers.14.self_attn.q_proj.weight": "
|
| 335 |
-
"layers.14.self_attn.q_proj.scales": "
|
| 336 |
-
"layers.14.self_attn.q_proj.biases": "
|
| 337 |
-
"layers.14.self_attn.k_proj.weight": "
|
| 338 |
-
"layers.14.self_attn.k_proj.scales": "
|
| 339 |
-
"layers.14.self_attn.k_proj.biases": "
|
| 340 |
-
"layers.14.self_attn.v_proj.weight": "
|
| 341 |
-
"layers.14.self_attn.v_proj.scales": "
|
| 342 |
-
"layers.14.self_attn.v_proj.biases": "
|
| 343 |
-
"layers.14.self_attn.o_proj.weight": "
|
| 344 |
-
"layers.14.self_attn.o_proj.scales": "
|
| 345 |
-
"layers.14.self_attn.o_proj.biases": "
|
| 346 |
-
"layers.14.mlp.gate_proj.weight": "
|
| 347 |
-
"layers.14.mlp.gate_proj.scales": "
|
| 348 |
-
"layers.14.mlp.gate_proj.biases": "
|
| 349 |
-
"layers.14.mlp.up_proj.weight": "
|
| 350 |
-
"layers.14.mlp.up_proj.scales": "
|
| 351 |
-
"layers.14.mlp.up_proj.biases": "
|
| 352 |
-
"layers.14.mlp.down_proj.weight": "
|
| 353 |
-
"layers.14.mlp.down_proj.scales": "
|
| 354 |
-
"layers.14.mlp.down_proj.biases": "
|
| 355 |
-
"layers.15.input_layernorm.weight": "
|
| 356 |
-
"layers.15.post_attention_layernorm.weight": "
|
| 357 |
-
"layers.15.self_attn.q_proj.weight": "
|
| 358 |
-
"layers.15.self_attn.q_proj.scales": "
|
| 359 |
-
"layers.15.self_attn.q_proj.biases": "
|
| 360 |
-
"layers.15.self_attn.k_proj.weight": "
|
| 361 |
-
"layers.15.self_attn.k_proj.scales": "
|
| 362 |
-
"layers.15.self_attn.k_proj.biases": "
|
| 363 |
-
"layers.15.self_attn.v_proj.weight": "
|
| 364 |
-
"layers.15.self_attn.v_proj.scales": "
|
| 365 |
-
"layers.15.self_attn.v_proj.biases": "
|
| 366 |
-
"layers.15.self_attn.o_proj.weight": "
|
| 367 |
-
"layers.15.self_attn.o_proj.scales": "
|
| 368 |
-
"layers.15.self_attn.o_proj.biases": "
|
| 369 |
-
"layers.15.mlp.gate_proj.weight": "
|
| 370 |
-
"layers.15.mlp.gate_proj.scales": "
|
| 371 |
-
"layers.15.mlp.gate_proj.biases": "
|
| 372 |
"layers.15.mlp.up_proj.weight": "1.safetensors",
|
| 373 |
"layers.15.mlp.up_proj.scales": "1.safetensors",
|
| 374 |
"layers.15.mlp.up_proj.biases": "1.safetensors",
|
|
|
|
| 1 |
{
|
| 2 |
"metadata": {
|
| 3 |
+
"quantization_level": "8",
|
| 4 |
"mflux_version": "0.18.5"
|
| 5 |
},
|
| 6 |
"weight_map": {
|
|
|
|
| 326 |
"layers.13.mlp.up_proj.weight": "0.safetensors",
|
| 327 |
"layers.13.mlp.up_proj.scales": "0.safetensors",
|
| 328 |
"layers.13.mlp.up_proj.biases": "0.safetensors",
|
| 329 |
+
"layers.13.mlp.down_proj.weight": "1.safetensors",
|
| 330 |
+
"layers.13.mlp.down_proj.scales": "1.safetensors",
|
| 331 |
+
"layers.13.mlp.down_proj.biases": "1.safetensors",
|
| 332 |
+
"layers.14.input_layernorm.weight": "1.safetensors",
|
| 333 |
+
"layers.14.post_attention_layernorm.weight": "1.safetensors",
|
| 334 |
+
"layers.14.self_attn.q_proj.weight": "1.safetensors",
|
| 335 |
+
"layers.14.self_attn.q_proj.scales": "1.safetensors",
|
| 336 |
+
"layers.14.self_attn.q_proj.biases": "1.safetensors",
|
| 337 |
+
"layers.14.self_attn.k_proj.weight": "1.safetensors",
|
| 338 |
+
"layers.14.self_attn.k_proj.scales": "1.safetensors",
|
| 339 |
+
"layers.14.self_attn.k_proj.biases": "1.safetensors",
|
| 340 |
+
"layers.14.self_attn.v_proj.weight": "1.safetensors",
|
| 341 |
+
"layers.14.self_attn.v_proj.scales": "1.safetensors",
|
| 342 |
+
"layers.14.self_attn.v_proj.biases": "1.safetensors",
|
| 343 |
+
"layers.14.self_attn.o_proj.weight": "1.safetensors",
|
| 344 |
+
"layers.14.self_attn.o_proj.scales": "1.safetensors",
|
| 345 |
+
"layers.14.self_attn.o_proj.biases": "1.safetensors",
|
| 346 |
+
"layers.14.mlp.gate_proj.weight": "1.safetensors",
|
| 347 |
+
"layers.14.mlp.gate_proj.scales": "1.safetensors",
|
| 348 |
+
"layers.14.mlp.gate_proj.biases": "1.safetensors",
|
| 349 |
+
"layers.14.mlp.up_proj.weight": "1.safetensors",
|
| 350 |
+
"layers.14.mlp.up_proj.scales": "1.safetensors",
|
| 351 |
+
"layers.14.mlp.up_proj.biases": "1.safetensors",
|
| 352 |
+
"layers.14.mlp.down_proj.weight": "1.safetensors",
|
| 353 |
+
"layers.14.mlp.down_proj.scales": "1.safetensors",
|
| 354 |
+
"layers.14.mlp.down_proj.biases": "1.safetensors",
|
| 355 |
+
"layers.15.input_layernorm.weight": "1.safetensors",
|
| 356 |
+
"layers.15.post_attention_layernorm.weight": "1.safetensors",
|
| 357 |
+
"layers.15.self_attn.q_proj.weight": "1.safetensors",
|
| 358 |
+
"layers.15.self_attn.q_proj.scales": "1.safetensors",
|
| 359 |
+
"layers.15.self_attn.q_proj.biases": "1.safetensors",
|
| 360 |
+
"layers.15.self_attn.k_proj.weight": "1.safetensors",
|
| 361 |
+
"layers.15.self_attn.k_proj.scales": "1.safetensors",
|
| 362 |
+
"layers.15.self_attn.k_proj.biases": "1.safetensors",
|
| 363 |
+
"layers.15.self_attn.v_proj.weight": "1.safetensors",
|
| 364 |
+
"layers.15.self_attn.v_proj.scales": "1.safetensors",
|
| 365 |
+
"layers.15.self_attn.v_proj.biases": "1.safetensors",
|
| 366 |
+
"layers.15.self_attn.o_proj.weight": "1.safetensors",
|
| 367 |
+
"layers.15.self_attn.o_proj.scales": "1.safetensors",
|
| 368 |
+
"layers.15.self_attn.o_proj.biases": "1.safetensors",
|
| 369 |
+
"layers.15.mlp.gate_proj.weight": "1.safetensors",
|
| 370 |
+
"layers.15.mlp.gate_proj.scales": "1.safetensors",
|
| 371 |
+
"layers.15.mlp.gate_proj.biases": "1.safetensors",
|
| 372 |
"layers.15.mlp.up_proj.weight": "1.safetensors",
|
| 373 |
"layers.15.mlp.up_proj.scales": "1.safetensors",
|
| 374 |
"layers.15.mlp.up_proj.biases": "1.safetensors",
|
transformer/0.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ecb1cf2d973c4ecbcfb0227a8c4bb8e328ce5d4bafafb19ae0bff07f5c12e1da
|
| 3 |
+
size 2135938597
|
transformer/1.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d13f396bdfedd2f70269a88d2821ebba3c1d3c007b73eb9710a2ddb3382feba
|
| 3 |
+
size 2139272193
|
transformer/2.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eded3f0f98b83f92fb784d765ea0871698ad97a6e3392cbd84a04f62467b6690
|
| 3 |
+
size 2139280460
|
transformer/3.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:feea529ed20503cefa38cc386bd00810919dedaa5d81e65b808e6dea019f8f74
|
| 3 |
+
size 2122012247
|
transformer/model.safetensors.index.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"metadata": {
|
| 3 |
-
"quantization_level": "
|
| 4 |
"mflux_version": "0.18.5"
|
| 5 |
},
|
| 6 |
"weight_map": {
|
|
@@ -240,149 +240,149 @@
|
|
| 240 |
"layers.8.mlp.gate_proj.weight": "0.safetensors",
|
| 241 |
"layers.8.mlp.gate_proj.scales": "0.safetensors",
|
| 242 |
"layers.8.mlp.gate_proj.biases": "0.safetensors",
|
| 243 |
-
"layers.8.mlp.up_proj.weight": "
|
| 244 |
-
"layers.8.mlp.up_proj.scales": "
|
| 245 |
-
"layers.8.mlp.up_proj.biases": "
|
| 246 |
-
"layers.8.mlp.linear_fc2.weight": "
|
| 247 |
-
"layers.8.mlp.linear_fc2.scales": "
|
| 248 |
-
"layers.8.mlp.linear_fc2.biases": "
|
| 249 |
-
"layers.9.adaLN_sa_ln.weight": "
|
| 250 |
-
"layers.9.self_attention.to_q.weight": "
|
| 251 |
-
"layers.9.self_attention.to_q.scales": "
|
| 252 |
-
"layers.9.self_attention.to_q.biases": "
|
| 253 |
-
"layers.9.self_attention.to_k.weight": "
|
| 254 |
-
"layers.9.self_attention.to_k.scales": "
|
| 255 |
-
"layers.9.self_attention.to_k.biases": "
|
| 256 |
-
"layers.9.self_attention.to_v.weight": "
|
| 257 |
-
"layers.9.self_attention.to_v.scales": "
|
| 258 |
-
"layers.9.self_attention.to_v.biases": "
|
| 259 |
-
"layers.9.self_attention.norm_q.weight": "
|
| 260 |
-
"layers.9.self_attention.norm_k.weight": "
|
| 261 |
-
"layers.9.self_attention.to_out.0.weight": "
|
| 262 |
-
"layers.9.self_attention.to_out.0.scales": "
|
| 263 |
-
"layers.9.self_attention.to_out.0.biases": "
|
| 264 |
-
"layers.9.adaLN_mlp_ln.weight": "
|
| 265 |
-
"layers.9.mlp.gate_proj.weight": "
|
| 266 |
-
"layers.9.mlp.gate_proj.scales": "
|
| 267 |
-
"layers.9.mlp.gate_proj.biases": "
|
| 268 |
-
"layers.9.mlp.up_proj.weight": "
|
| 269 |
-
"layers.9.mlp.up_proj.scales": "
|
| 270 |
-
"layers.9.mlp.up_proj.biases": "
|
| 271 |
-
"layers.9.mlp.linear_fc2.weight": "
|
| 272 |
-
"layers.9.mlp.linear_fc2.scales": "
|
| 273 |
-
"layers.9.mlp.linear_fc2.biases": "
|
| 274 |
-
"layers.10.adaLN_sa_ln.weight": "
|
| 275 |
-
"layers.10.self_attention.to_q.weight": "
|
| 276 |
-
"layers.10.self_attention.to_q.scales": "
|
| 277 |
-
"layers.10.self_attention.to_q.biases": "
|
| 278 |
-
"layers.10.self_attention.to_k.weight": "
|
| 279 |
-
"layers.10.self_attention.to_k.scales": "
|
| 280 |
-
"layers.10.self_attention.to_k.biases": "
|
| 281 |
-
"layers.10.self_attention.to_v.weight": "
|
| 282 |
-
"layers.10.self_attention.to_v.scales": "
|
| 283 |
-
"layers.10.self_attention.to_v.biases": "
|
| 284 |
-
"layers.10.self_attention.norm_q.weight": "
|
| 285 |
-
"layers.10.self_attention.norm_k.weight": "
|
| 286 |
-
"layers.10.self_attention.to_out.0.weight": "
|
| 287 |
-
"layers.10.self_attention.to_out.0.scales": "
|
| 288 |
-
"layers.10.self_attention.to_out.0.biases": "
|
| 289 |
-
"layers.10.adaLN_mlp_ln.weight": "
|
| 290 |
-
"layers.10.mlp.gate_proj.weight": "
|
| 291 |
-
"layers.10.mlp.gate_proj.scales": "
|
| 292 |
-
"layers.10.mlp.gate_proj.biases": "
|
| 293 |
-
"layers.10.mlp.up_proj.weight": "
|
| 294 |
-
"layers.10.mlp.up_proj.scales": "
|
| 295 |
-
"layers.10.mlp.up_proj.biases": "
|
| 296 |
-
"layers.10.mlp.linear_fc2.weight": "
|
| 297 |
-
"layers.10.mlp.linear_fc2.scales": "
|
| 298 |
-
"layers.10.mlp.linear_fc2.biases": "
|
| 299 |
-
"layers.11.adaLN_sa_ln.weight": "
|
| 300 |
-
"layers.11.self_attention.to_q.weight": "
|
| 301 |
-
"layers.11.self_attention.to_q.scales": "
|
| 302 |
-
"layers.11.self_attention.to_q.biases": "
|
| 303 |
-
"layers.11.self_attention.to_k.weight": "
|
| 304 |
-
"layers.11.self_attention.to_k.scales": "
|
| 305 |
-
"layers.11.self_attention.to_k.biases": "
|
| 306 |
-
"layers.11.self_attention.to_v.weight": "
|
| 307 |
-
"layers.11.self_attention.to_v.scales": "
|
| 308 |
-
"layers.11.self_attention.to_v.biases": "
|
| 309 |
-
"layers.11.self_attention.norm_q.weight": "
|
| 310 |
-
"layers.11.self_attention.norm_k.weight": "
|
| 311 |
-
"layers.11.self_attention.to_out.0.weight": "
|
| 312 |
-
"layers.11.self_attention.to_out.0.scales": "
|
| 313 |
-
"layers.11.self_attention.to_out.0.biases": "
|
| 314 |
-
"layers.11.adaLN_mlp_ln.weight": "
|
| 315 |
-
"layers.11.mlp.gate_proj.weight": "
|
| 316 |
-
"layers.11.mlp.gate_proj.scales": "
|
| 317 |
-
"layers.11.mlp.gate_proj.biases": "
|
| 318 |
-
"layers.11.mlp.up_proj.weight": "
|
| 319 |
-
"layers.11.mlp.up_proj.scales": "
|
| 320 |
-
"layers.11.mlp.up_proj.biases": "
|
| 321 |
-
"layers.11.mlp.linear_fc2.weight": "
|
| 322 |
-
"layers.11.mlp.linear_fc2.scales": "
|
| 323 |
-
"layers.11.mlp.linear_fc2.biases": "
|
| 324 |
-
"layers.12.adaLN_sa_ln.weight": "
|
| 325 |
-
"layers.12.self_attention.to_q.weight": "
|
| 326 |
-
"layers.12.self_attention.to_q.scales": "
|
| 327 |
-
"layers.12.self_attention.to_q.biases": "
|
| 328 |
-
"layers.12.self_attention.to_k.weight": "
|
| 329 |
-
"layers.12.self_attention.to_k.scales": "
|
| 330 |
-
"layers.12.self_attention.to_k.biases": "
|
| 331 |
-
"layers.12.self_attention.to_v.weight": "
|
| 332 |
-
"layers.12.self_attention.to_v.scales": "
|
| 333 |
-
"layers.12.self_attention.to_v.biases": "
|
| 334 |
-
"layers.12.self_attention.norm_q.weight": "
|
| 335 |
-
"layers.12.self_attention.norm_k.weight": "
|
| 336 |
-
"layers.12.self_attention.to_out.0.weight": "
|
| 337 |
-
"layers.12.self_attention.to_out.0.scales": "
|
| 338 |
-
"layers.12.self_attention.to_out.0.biases": "
|
| 339 |
-
"layers.12.adaLN_mlp_ln.weight": "
|
| 340 |
-
"layers.12.mlp.gate_proj.weight": "
|
| 341 |
-
"layers.12.mlp.gate_proj.scales": "
|
| 342 |
-
"layers.12.mlp.gate_proj.biases": "
|
| 343 |
-
"layers.12.mlp.up_proj.weight": "
|
| 344 |
-
"layers.12.mlp.up_proj.scales": "
|
| 345 |
-
"layers.12.mlp.up_proj.biases": "
|
| 346 |
-
"layers.12.mlp.linear_fc2.weight": "
|
| 347 |
-
"layers.12.mlp.linear_fc2.scales": "
|
| 348 |
-
"layers.12.mlp.linear_fc2.biases": "
|
| 349 |
-
"layers.13.adaLN_sa_ln.weight": "
|
| 350 |
-
"layers.13.self_attention.to_q.weight": "
|
| 351 |
-
"layers.13.self_attention.to_q.scales": "
|
| 352 |
-
"layers.13.self_attention.to_q.biases": "
|
| 353 |
-
"layers.13.self_attention.to_k.weight": "
|
| 354 |
-
"layers.13.self_attention.to_k.scales": "
|
| 355 |
-
"layers.13.self_attention.to_k.biases": "
|
| 356 |
-
"layers.13.self_attention.to_v.weight": "
|
| 357 |
-
"layers.13.self_attention.to_v.scales": "
|
| 358 |
-
"layers.13.self_attention.to_v.biases": "
|
| 359 |
-
"layers.13.self_attention.norm_q.weight": "
|
| 360 |
-
"layers.13.self_attention.norm_k.weight": "
|
| 361 |
-
"layers.13.self_attention.to_out.0.weight": "
|
| 362 |
-
"layers.13.self_attention.to_out.0.scales": "
|
| 363 |
-
"layers.13.self_attention.to_out.0.biases": "
|
| 364 |
-
"layers.13.adaLN_mlp_ln.weight": "
|
| 365 |
-
"layers.13.mlp.gate_proj.weight": "
|
| 366 |
-
"layers.13.mlp.gate_proj.scales": "
|
| 367 |
-
"layers.13.mlp.gate_proj.biases": "
|
| 368 |
-
"layers.13.mlp.up_proj.weight": "
|
| 369 |
-
"layers.13.mlp.up_proj.scales": "
|
| 370 |
-
"layers.13.mlp.up_proj.biases": "
|
| 371 |
-
"layers.13.mlp.linear_fc2.weight": "
|
| 372 |
-
"layers.13.mlp.linear_fc2.scales": "
|
| 373 |
-
"layers.13.mlp.linear_fc2.biases": "
|
| 374 |
-
"layers.14.adaLN_sa_ln.weight": "
|
| 375 |
-
"layers.14.self_attention.to_q.weight": "
|
| 376 |
-
"layers.14.self_attention.to_q.scales": "
|
| 377 |
-
"layers.14.self_attention.to_q.biases": "
|
| 378 |
-
"layers.14.self_attention.to_k.weight": "
|
| 379 |
-
"layers.14.self_attention.to_k.scales": "
|
| 380 |
-
"layers.14.self_attention.to_k.biases": "
|
| 381 |
-
"layers.14.self_attention.to_v.weight": "
|
| 382 |
-
"layers.14.self_attention.to_v.scales": "
|
| 383 |
-
"layers.14.self_attention.to_v.biases": "
|
| 384 |
-
"layers.14.self_attention.norm_q.weight": "
|
| 385 |
-
"layers.14.self_attention.norm_k.weight": "
|
| 386 |
"layers.14.self_attention.to_out.0.weight": "1.safetensors",
|
| 387 |
"layers.14.self_attention.to_out.0.scales": "1.safetensors",
|
| 388 |
"layers.14.self_attention.to_out.0.biases": "1.safetensors",
|
|
@@ -468,466 +468,466 @@
|
|
| 468 |
"layers.17.mlp.up_proj.weight": "1.safetensors",
|
| 469 |
"layers.17.mlp.up_proj.scales": "1.safetensors",
|
| 470 |
"layers.17.mlp.up_proj.biases": "1.safetensors",
|
| 471 |
-
"layers.17.mlp.linear_fc2.weight": "
|
| 472 |
-
"layers.17.mlp.linear_fc2.scales": "
|
| 473 |
-
"layers.17.mlp.linear_fc2.biases": "
|
| 474 |
-
"layers.18.adaLN_sa_ln.weight": "
|
| 475 |
-
"layers.18.self_attention.to_q.weight": "
|
| 476 |
-
"layers.18.self_attention.to_q.scales": "
|
| 477 |
-
"layers.18.self_attention.to_q.biases": "
|
| 478 |
-
"layers.18.self_attention.to_k.weight": "
|
| 479 |
-
"layers.18.self_attention.to_k.scales": "
|
| 480 |
-
"layers.18.self_attention.to_k.biases": "
|
| 481 |
-
"layers.18.self_attention.to_v.weight": "
|
| 482 |
-
"layers.18.self_attention.to_v.scales": "
|
| 483 |
-
"layers.18.self_attention.to_v.biases": "
|
| 484 |
-
"layers.18.self_attention.norm_q.weight": "
|
| 485 |
-
"layers.18.self_attention.norm_k.weight": "
|
| 486 |
-
"layers.18.self_attention.to_out.0.weight": "
|
| 487 |
-
"layers.18.self_attention.to_out.0.scales": "
|
| 488 |
-
"layers.18.self_attention.to_out.0.biases": "
|
| 489 |
-
"layers.18.adaLN_mlp_ln.weight": "
|
| 490 |
-
"layers.18.mlp.gate_proj.weight": "
|
| 491 |
-
"layers.18.mlp.gate_proj.scales": "
|
| 492 |
-
"layers.18.mlp.gate_proj.biases": "
|
| 493 |
-
"layers.18.mlp.up_proj.weight": "
|
| 494 |
-
"layers.18.mlp.up_proj.scales": "
|
| 495 |
-
"layers.18.mlp.up_proj.biases": "
|
| 496 |
-
"layers.18.mlp.linear_fc2.weight": "
|
| 497 |
-
"layers.18.mlp.linear_fc2.scales": "
|
| 498 |
-
"layers.18.mlp.linear_fc2.biases": "
|
| 499 |
-
"layers.19.adaLN_sa_ln.weight": "
|
| 500 |
-
"layers.19.self_attention.to_q.weight": "
|
| 501 |
-
"layers.19.self_attention.to_q.scales": "
|
| 502 |
-
"layers.19.self_attention.to_q.biases": "
|
| 503 |
-
"layers.19.self_attention.to_k.weight": "
|
| 504 |
-
"layers.19.self_attention.to_k.scales": "
|
| 505 |
-
"layers.19.self_attention.to_k.biases": "
|
| 506 |
-
"layers.19.self_attention.to_v.weight": "
|
| 507 |
-
"layers.19.self_attention.to_v.scales": "
|
| 508 |
-
"layers.19.self_attention.to_v.biases": "
|
| 509 |
-
"layers.19.self_attention.norm_q.weight": "
|
| 510 |
-
"layers.19.self_attention.norm_k.weight": "
|
| 511 |
-
"layers.19.self_attention.to_out.0.weight": "
|
| 512 |
-
"layers.19.self_attention.to_out.0.scales": "
|
| 513 |
-
"layers.19.self_attention.to_out.0.biases": "
|
| 514 |
-
"layers.19.adaLN_mlp_ln.weight": "
|
| 515 |
-
"layers.19.mlp.gate_proj.weight": "
|
| 516 |
-
"layers.19.mlp.gate_proj.scales": "
|
| 517 |
-
"layers.19.mlp.gate_proj.biases": "
|
| 518 |
-
"layers.19.mlp.up_proj.weight": "
|
| 519 |
-
"layers.19.mlp.up_proj.scales": "
|
| 520 |
-
"layers.19.mlp.up_proj.biases": "
|
| 521 |
-
"layers.19.mlp.linear_fc2.weight": "
|
| 522 |
-
"layers.19.mlp.linear_fc2.scales": "
|
| 523 |
-
"layers.19.mlp.linear_fc2.biases": "
|
| 524 |
-
"layers.20.adaLN_sa_ln.weight": "
|
| 525 |
-
"layers.20.self_attention.to_q.weight": "
|
| 526 |
-
"layers.20.self_attention.to_q.scales": "
|
| 527 |
-
"layers.20.self_attention.to_q.biases": "
|
| 528 |
-
"layers.20.self_attention.to_k.weight": "
|
| 529 |
-
"layers.20.self_attention.to_k.scales": "
|
| 530 |
-
"layers.20.self_attention.to_k.biases": "
|
| 531 |
-
"layers.20.self_attention.to_v.weight": "
|
| 532 |
-
"layers.20.self_attention.to_v.scales": "
|
| 533 |
-
"layers.20.self_attention.to_v.biases": "
|
| 534 |
-
"layers.20.self_attention.norm_q.weight": "
|
| 535 |
-
"layers.20.self_attention.norm_k.weight": "
|
| 536 |
-
"layers.20.self_attention.to_out.0.weight": "
|
| 537 |
-
"layers.20.self_attention.to_out.0.scales": "
|
| 538 |
-
"layers.20.self_attention.to_out.0.biases": "
|
| 539 |
-
"layers.20.adaLN_mlp_ln.weight": "
|
| 540 |
-
"layers.20.mlp.gate_proj.weight": "
|
| 541 |
-
"layers.20.mlp.gate_proj.scales": "
|
| 542 |
-
"layers.20.mlp.gate_proj.biases": "
|
| 543 |
-
"layers.20.mlp.up_proj.weight": "
|
| 544 |
-
"layers.20.mlp.up_proj.scales": "
|
| 545 |
-
"layers.20.mlp.up_proj.biases": "
|
| 546 |
-
"layers.20.mlp.linear_fc2.weight": "
|
| 547 |
-
"layers.20.mlp.linear_fc2.scales": "
|
| 548 |
-
"layers.20.mlp.linear_fc2.biases": "
|
| 549 |
-
"layers.21.adaLN_sa_ln.weight": "
|
| 550 |
-
"layers.21.self_attention.to_q.weight": "
|
| 551 |
-
"layers.21.self_attention.to_q.scales": "
|
| 552 |
-
"layers.21.self_attention.to_q.biases": "
|
| 553 |
-
"layers.21.self_attention.to_k.weight": "
|
| 554 |
-
"layers.21.self_attention.to_k.scales": "
|
| 555 |
-
"layers.21.self_attention.to_k.biases": "
|
| 556 |
-
"layers.21.self_attention.to_v.weight": "
|
| 557 |
-
"layers.21.self_attention.to_v.scales": "
|
| 558 |
-
"layers.21.self_attention.to_v.biases": "
|
| 559 |
-
"layers.21.self_attention.norm_q.weight": "
|
| 560 |
-
"layers.21.self_attention.norm_k.weight": "
|
| 561 |
-
"layers.21.self_attention.to_out.0.weight": "
|
| 562 |
-
"layers.21.self_attention.to_out.0.scales": "
|
| 563 |
-
"layers.21.self_attention.to_out.0.biases": "
|
| 564 |
-
"layers.21.adaLN_mlp_ln.weight": "
|
| 565 |
-
"layers.21.mlp.gate_proj.weight": "
|
| 566 |
-
"layers.21.mlp.gate_proj.scales": "
|
| 567 |
-
"layers.21.mlp.gate_proj.biases": "
|
| 568 |
-
"layers.21.mlp.up_proj.weight": "
|
| 569 |
-
"layers.21.mlp.up_proj.scales": "
|
| 570 |
-
"layers.21.mlp.up_proj.biases": "
|
| 571 |
-
"layers.21.mlp.linear_fc2.weight": "
|
| 572 |
-
"layers.21.mlp.linear_fc2.scales": "
|
| 573 |
-
"layers.21.mlp.linear_fc2.biases": "
|
| 574 |
-
"layers.22.adaLN_sa_ln.weight": "
|
| 575 |
-
"layers.22.self_attention.to_q.weight": "
|
| 576 |
-
"layers.22.self_attention.to_q.scales": "
|
| 577 |
-
"layers.22.self_attention.to_q.biases": "
|
| 578 |
-
"layers.22.self_attention.to_k.weight": "
|
| 579 |
-
"layers.22.self_attention.to_k.scales": "
|
| 580 |
-
"layers.22.self_attention.to_k.biases": "
|
| 581 |
-
"layers.22.self_attention.to_v.weight": "
|
| 582 |
-
"layers.22.self_attention.to_v.scales": "
|
| 583 |
-
"layers.22.self_attention.to_v.biases": "
|
| 584 |
-
"layers.22.self_attention.norm_q.weight": "
|
| 585 |
-
"layers.22.self_attention.norm_k.weight": "
|
| 586 |
-
"layers.22.self_attention.to_out.0.weight": "
|
| 587 |
-
"layers.22.self_attention.to_out.0.scales": "
|
| 588 |
-
"layers.22.self_attention.to_out.0.biases": "
|
| 589 |
-
"layers.22.adaLN_mlp_ln.weight": "
|
| 590 |
-
"layers.22.mlp.gate_proj.weight": "
|
| 591 |
-
"layers.22.mlp.gate_proj.scales": "
|
| 592 |
-
"layers.22.mlp.gate_proj.biases": "
|
| 593 |
-
"layers.22.mlp.up_proj.weight": "
|
| 594 |
-
"layers.22.mlp.up_proj.scales": "
|
| 595 |
-
"layers.22.mlp.up_proj.biases": "
|
| 596 |
-
"layers.22.mlp.linear_fc2.weight": "
|
| 597 |
-
"layers.22.mlp.linear_fc2.scales": "
|
| 598 |
-
"layers.22.mlp.linear_fc2.biases": "
|
| 599 |
-
"layers.23.adaLN_sa_ln.weight": "
|
| 600 |
-
"layers.23.self_attention.to_q.weight": "
|
| 601 |
-
"layers.23.self_attention.to_q.scales": "
|
| 602 |
-
"layers.23.self_attention.to_q.biases": "
|
| 603 |
-
"layers.23.self_attention.to_k.weight": "
|
| 604 |
-
"layers.23.self_attention.to_k.scales": "
|
| 605 |
-
"layers.23.self_attention.to_k.biases": "
|
| 606 |
-
"layers.23.self_attention.to_v.weight": "
|
| 607 |
-
"layers.23.self_attention.to_v.scales": "
|
| 608 |
-
"layers.23.self_attention.to_v.biases": "
|
| 609 |
-
"layers.23.self_attention.norm_q.weight": "
|
| 610 |
-
"layers.23.self_attention.norm_k.weight": "
|
| 611 |
-
"layers.23.self_attention.to_out.0.weight": "
|
| 612 |
-
"layers.23.self_attention.to_out.0.scales": "
|
| 613 |
-
"layers.23.self_attention.to_out.0.biases": "
|
| 614 |
-
"layers.23.adaLN_mlp_ln.weight": "
|
| 615 |
-
"layers.23.mlp.gate_proj.weight": "
|
| 616 |
-
"layers.23.mlp.gate_proj.scales": "
|
| 617 |
-
"layers.23.mlp.gate_proj.biases": "
|
| 618 |
-
"layers.23.mlp.up_proj.weight": "
|
| 619 |
-
"layers.23.mlp.up_proj.scales": "
|
| 620 |
-
"layers.23.mlp.up_proj.biases": "
|
| 621 |
-
"layers.23.mlp.linear_fc2.weight": "
|
| 622 |
-
"layers.23.mlp.linear_fc2.scales": "
|
| 623 |
-
"layers.23.mlp.linear_fc2.biases": "
|
| 624 |
-
"layers.24.adaLN_sa_ln.weight": "
|
| 625 |
-
"layers.24.self_attention.to_q.weight": "
|
| 626 |
-
"layers.24.self_attention.to_q.scales": "
|
| 627 |
-
"layers.24.self_attention.to_q.biases": "
|
| 628 |
-
"layers.24.self_attention.to_k.weight": "
|
| 629 |
-
"layers.24.self_attention.to_k.scales": "
|
| 630 |
-
"layers.24.self_attention.to_k.biases": "
|
| 631 |
-
"layers.24.self_attention.to_v.weight": "
|
| 632 |
-
"layers.24.self_attention.to_v.scales": "
|
| 633 |
-
"layers.24.self_attention.to_v.biases": "
|
| 634 |
-
"layers.24.self_attention.norm_q.weight": "
|
| 635 |
-
"layers.24.self_attention.norm_k.weight": "
|
| 636 |
-
"layers.24.self_attention.to_out.0.weight": "
|
| 637 |
-
"layers.24.self_attention.to_out.0.scales": "
|
| 638 |
-
"layers.24.self_attention.to_out.0.biases": "
|
| 639 |
-
"layers.24.adaLN_mlp_ln.weight": "
|
| 640 |
-
"layers.24.mlp.gate_proj.weight": "
|
| 641 |
-
"layers.24.mlp.gate_proj.scales": "
|
| 642 |
-
"layers.24.mlp.gate_proj.biases": "
|
| 643 |
-
"layers.24.mlp.up_proj.weight": "
|
| 644 |
-
"layers.24.mlp.up_proj.scales": "
|
| 645 |
-
"layers.24.mlp.up_proj.biases": "
|
| 646 |
-
"layers.24.mlp.linear_fc2.weight": "
|
| 647 |
-
"layers.24.mlp.linear_fc2.scales": "
|
| 648 |
-
"layers.24.mlp.linear_fc2.biases": "
|
| 649 |
-
"layers.25.adaLN_sa_ln.weight": "
|
| 650 |
-
"layers.25.self_attention.to_q.weight": "
|
| 651 |
-
"layers.25.self_attention.to_q.scales": "
|
| 652 |
-
"layers.25.self_attention.to_q.biases": "
|
| 653 |
-
"layers.25.self_attention.to_k.weight": "
|
| 654 |
-
"layers.25.self_attention.to_k.scales": "
|
| 655 |
-
"layers.25.self_attention.to_k.biases": "
|
| 656 |
-
"layers.25.self_attention.to_v.weight": "
|
| 657 |
-
"layers.25.self_attention.to_v.scales": "
|
| 658 |
-
"layers.25.self_attention.to_v.biases": "
|
| 659 |
-
"layers.25.self_attention.norm_q.weight": "
|
| 660 |
-
"layers.25.self_attention.norm_k.weight": "
|
| 661 |
-
"layers.25.self_attention.to_out.0.weight": "
|
| 662 |
-
"layers.25.self_attention.to_out.0.scales": "
|
| 663 |
-
"layers.25.self_attention.to_out.0.biases": "
|
| 664 |
-
"layers.25.adaLN_mlp_ln.weight": "
|
| 665 |
-
"layers.25.mlp.gate_proj.weight": "
|
| 666 |
-
"layers.25.mlp.gate_proj.scales": "
|
| 667 |
-
"layers.25.mlp.gate_proj.biases": "
|
| 668 |
-
"layers.25.mlp.up_proj.weight": "
|
| 669 |
-
"layers.25.mlp.up_proj.scales": "
|
| 670 |
-
"layers.25.mlp.up_proj.biases": "
|
| 671 |
-
"layers.25.mlp.linear_fc2.weight": "
|
| 672 |
-
"layers.25.mlp.linear_fc2.scales": "
|
| 673 |
-
"layers.25.mlp.linear_fc2.biases": "
|
| 674 |
-
"layers.26.adaLN_sa_ln.weight": "
|
| 675 |
-
"layers.26.self_attention.to_q.weight": "
|
| 676 |
-
"layers.26.self_attention.to_q.scales": "
|
| 677 |
-
"layers.26.self_attention.to_q.biases": "
|
| 678 |
-
"layers.26.self_attention.to_k.weight": "
|
| 679 |
-
"layers.26.self_attention.to_k.scales": "
|
| 680 |
-
"layers.26.self_attention.to_k.biases": "
|
| 681 |
-
"layers.26.self_attention.to_v.weight": "
|
| 682 |
-
"layers.26.self_attention.to_v.scales": "
|
| 683 |
-
"layers.26.self_attention.to_v.biases": "
|
| 684 |
-
"layers.26.self_attention.norm_q.weight": "
|
| 685 |
-
"layers.26.self_attention.norm_k.weight": "
|
| 686 |
-
"layers.26.self_attention.to_out.0.weight": "
|
| 687 |
-
"layers.26.self_attention.to_out.0.scales": "
|
| 688 |
-
"layers.26.self_attention.to_out.0.biases": "
|
| 689 |
-
"layers.26.adaLN_mlp_ln.weight": "
|
| 690 |
-
"layers.26.mlp.gate_proj.weight": "
|
| 691 |
-
"layers.26.mlp.gate_proj.scales": "
|
| 692 |
-
"layers.26.mlp.gate_proj.biases": "
|
| 693 |
-
"layers.26.mlp.up_proj.weight": "
|
| 694 |
-
"layers.26.mlp.up_proj.scales": "
|
| 695 |
-
"layers.26.mlp.up_proj.biases": "
|
| 696 |
-
"layers.26.mlp.linear_fc2.weight": "
|
| 697 |
-
"layers.26.mlp.linear_fc2.scales": "
|
| 698 |
-
"layers.26.mlp.linear_fc2.biases": "
|
| 699 |
-
"layers.27.adaLN_sa_ln.weight": "
|
| 700 |
-
"layers.27.self_attention.to_q.weight": "
|
| 701 |
-
"layers.27.self_attention.to_q.scales": "
|
| 702 |
-
"layers.27.self_attention.to_q.biases": "
|
| 703 |
-
"layers.27.self_attention.to_k.weight": "
|
| 704 |
-
"layers.27.self_attention.to_k.scales": "
|
| 705 |
-
"layers.27.self_attention.to_k.biases": "
|
| 706 |
-
"layers.27.self_attention.to_v.weight": "
|
| 707 |
-
"layers.27.self_attention.to_v.scales": "
|
| 708 |
-
"layers.27.self_attention.to_v.biases": "
|
| 709 |
-
"layers.27.self_attention.norm_q.weight": "
|
| 710 |
-
"layers.27.self_attention.norm_k.weight": "
|
| 711 |
-
"layers.27.self_attention.to_out.0.weight": "
|
| 712 |
-
"layers.27.self_attention.to_out.0.scales": "
|
| 713 |
-
"layers.27.self_attention.to_out.0.biases": "
|
| 714 |
-
"layers.27.adaLN_mlp_ln.weight": "
|
| 715 |
-
"layers.27.mlp.gate_proj.weight": "
|
| 716 |
-
"layers.27.mlp.gate_proj.scales": "
|
| 717 |
-
"layers.27.mlp.gate_proj.biases": "
|
| 718 |
-
"layers.27.mlp.up_proj.weight": "
|
| 719 |
-
"layers.27.mlp.up_proj.scales": "
|
| 720 |
-
"layers.27.mlp.up_proj.biases": "
|
| 721 |
-
"layers.27.mlp.linear_fc2.weight": "
|
| 722 |
-
"layers.27.mlp.linear_fc2.scales": "
|
| 723 |
-
"layers.27.mlp.linear_fc2.biases": "
|
| 724 |
-
"layers.28.adaLN_sa_ln.weight": "
|
| 725 |
-
"layers.28.self_attention.to_q.weight": "
|
| 726 |
-
"layers.28.self_attention.to_q.scales": "
|
| 727 |
-
"layers.28.self_attention.to_q.biases": "
|
| 728 |
-
"layers.28.self_attention.to_k.weight": "
|
| 729 |
-
"layers.28.self_attention.to_k.scales": "
|
| 730 |
-
"layers.28.self_attention.to_k.biases": "
|
| 731 |
-
"layers.28.self_attention.to_v.weight": "
|
| 732 |
-
"layers.28.self_attention.to_v.scales": "
|
| 733 |
-
"layers.28.self_attention.to_v.biases": "
|
| 734 |
-
"layers.28.self_attention.norm_q.weight": "
|
| 735 |
-
"layers.28.self_attention.norm_k.weight": "
|
| 736 |
-
"layers.28.self_attention.to_out.0.weight": "
|
| 737 |
-
"layers.28.self_attention.to_out.0.scales": "
|
| 738 |
-
"layers.28.self_attention.to_out.0.biases": "
|
| 739 |
-
"layers.28.adaLN_mlp_ln.weight": "
|
| 740 |
-
"layers.28.mlp.gate_proj.weight": "
|
| 741 |
-
"layers.28.mlp.gate_proj.scales": "
|
| 742 |
-
"layers.28.mlp.gate_proj.biases": "
|
| 743 |
-
"layers.28.mlp.up_proj.weight": "
|
| 744 |
-
"layers.28.mlp.up_proj.scales": "
|
| 745 |
-
"layers.28.mlp.up_proj.biases": "
|
| 746 |
-
"layers.28.mlp.linear_fc2.weight": "
|
| 747 |
-
"layers.28.mlp.linear_fc2.scales": "
|
| 748 |
-
"layers.28.mlp.linear_fc2.biases": "
|
| 749 |
-
"layers.29.adaLN_sa_ln.weight": "
|
| 750 |
-
"layers.29.self_attention.to_q.weight": "
|
| 751 |
-
"layers.29.self_attention.to_q.scales": "
|
| 752 |
-
"layers.29.self_attention.to_q.biases": "
|
| 753 |
-
"layers.29.self_attention.to_k.weight": "
|
| 754 |
-
"layers.29.self_attention.to_k.scales": "
|
| 755 |
-
"layers.29.self_attention.to_k.biases": "
|
| 756 |
-
"layers.29.self_attention.to_v.weight": "
|
| 757 |
-
"layers.29.self_attention.to_v.scales": "
|
| 758 |
-
"layers.29.self_attention.to_v.biases": "
|
| 759 |
-
"layers.29.self_attention.norm_q.weight": "
|
| 760 |
-
"layers.29.self_attention.norm_k.weight": "
|
| 761 |
-
"layers.29.self_attention.to_out.0.weight": "
|
| 762 |
-
"layers.29.self_attention.to_out.0.scales": "
|
| 763 |
-
"layers.29.self_attention.to_out.0.biases": "
|
| 764 |
-
"layers.29.adaLN_mlp_ln.weight": "
|
| 765 |
-
"layers.29.mlp.gate_proj.weight": "
|
| 766 |
-
"layers.29.mlp.gate_proj.scales": "
|
| 767 |
-
"layers.29.mlp.gate_proj.biases": "
|
| 768 |
-
"layers.29.mlp.up_proj.weight": "
|
| 769 |
-
"layers.29.mlp.up_proj.scales": "
|
| 770 |
-
"layers.29.mlp.up_proj.biases": "
|
| 771 |
-
"layers.29.mlp.linear_fc2.weight": "
|
| 772 |
-
"layers.29.mlp.linear_fc2.scales": "
|
| 773 |
-
"layers.29.mlp.linear_fc2.biases": "
|
| 774 |
-
"layers.30.adaLN_sa_ln.weight": "
|
| 775 |
-
"layers.30.self_attention.to_q.weight": "
|
| 776 |
-
"layers.30.self_attention.to_q.scales": "
|
| 777 |
-
"layers.30.self_attention.to_q.biases": "
|
| 778 |
-
"layers.30.self_attention.to_k.weight": "
|
| 779 |
-
"layers.30.self_attention.to_k.scales": "
|
| 780 |
-
"layers.30.self_attention.to_k.biases": "
|
| 781 |
-
"layers.30.self_attention.to_v.weight": "
|
| 782 |
-
"layers.30.self_attention.to_v.scales": "
|
| 783 |
-
"layers.30.self_attention.to_v.biases": "
|
| 784 |
-
"layers.30.self_attention.norm_q.weight": "
|
| 785 |
-
"layers.30.self_attention.norm_k.weight": "
|
| 786 |
-
"layers.30.self_attention.to_out.0.weight": "
|
| 787 |
-
"layers.30.self_attention.to_out.0.scales": "
|
| 788 |
-
"layers.30.self_attention.to_out.0.biases": "
|
| 789 |
-
"layers.30.adaLN_mlp_ln.weight": "
|
| 790 |
-
"layers.30.mlp.gate_proj.weight": "
|
| 791 |
-
"layers.30.mlp.gate_proj.scales": "
|
| 792 |
-
"layers.30.mlp.gate_proj.biases": "
|
| 793 |
-
"layers.30.mlp.up_proj.weight": "
|
| 794 |
-
"layers.30.mlp.up_proj.scales": "
|
| 795 |
-
"layers.30.mlp.up_proj.biases": "
|
| 796 |
-
"layers.30.mlp.linear_fc2.weight": "
|
| 797 |
-
"layers.30.mlp.linear_fc2.scales": "
|
| 798 |
-
"layers.30.mlp.linear_fc2.biases": "
|
| 799 |
-
"layers.31.adaLN_sa_ln.weight": "
|
| 800 |
-
"layers.31.self_attention.to_q.weight": "
|
| 801 |
-
"layers.31.self_attention.to_q.scales": "
|
| 802 |
-
"layers.31.self_attention.to_q.biases": "
|
| 803 |
-
"layers.31.self_attention.to_k.weight": "
|
| 804 |
-
"layers.31.self_attention.to_k.scales": "
|
| 805 |
-
"layers.31.self_attention.to_k.biases": "
|
| 806 |
-
"layers.31.self_attention.to_v.weight": "
|
| 807 |
-
"layers.31.self_attention.to_v.scales": "
|
| 808 |
-
"layers.31.self_attention.to_v.biases": "
|
| 809 |
-
"layers.31.self_attention.norm_q.weight": "
|
| 810 |
-
"layers.31.self_attention.norm_k.weight": "
|
| 811 |
-
"layers.31.self_attention.to_out.0.weight": "
|
| 812 |
-
"layers.31.self_attention.to_out.0.scales": "
|
| 813 |
-
"layers.31.self_attention.to_out.0.biases": "
|
| 814 |
-
"layers.31.adaLN_mlp_ln.weight": "
|
| 815 |
-
"layers.31.mlp.gate_proj.weight": "
|
| 816 |
-
"layers.31.mlp.gate_proj.scales": "
|
| 817 |
-
"layers.31.mlp.gate_proj.biases": "
|
| 818 |
-
"layers.31.mlp.up_proj.weight": "
|
| 819 |
-
"layers.31.mlp.up_proj.scales": "
|
| 820 |
-
"layers.31.mlp.up_proj.biases": "
|
| 821 |
-
"layers.31.mlp.linear_fc2.weight": "
|
| 822 |
-
"layers.31.mlp.linear_fc2.scales": "
|
| 823 |
-
"layers.31.mlp.linear_fc2.biases": "
|
| 824 |
-
"layers.32.adaLN_sa_ln.weight": "
|
| 825 |
-
"layers.32.self_attention.to_q.weight": "
|
| 826 |
-
"layers.32.self_attention.to_q.scales": "
|
| 827 |
-
"layers.32.self_attention.to_q.biases": "
|
| 828 |
-
"layers.32.self_attention.to_k.weight": "
|
| 829 |
-
"layers.32.self_attention.to_k.scales": "
|
| 830 |
-
"layers.32.self_attention.to_k.biases": "
|
| 831 |
-
"layers.32.self_attention.to_v.weight": "
|
| 832 |
-
"layers.32.self_attention.to_v.scales": "
|
| 833 |
-
"layers.32.self_attention.to_v.biases": "
|
| 834 |
-
"layers.32.self_attention.norm_q.weight": "
|
| 835 |
-
"layers.32.self_attention.norm_k.weight": "
|
| 836 |
-
"layers.32.self_attention.to_out.0.weight": "
|
| 837 |
-
"layers.32.self_attention.to_out.0.scales": "
|
| 838 |
-
"layers.32.self_attention.to_out.0.biases": "
|
| 839 |
-
"layers.32.adaLN_mlp_ln.weight": "
|
| 840 |
-
"layers.32.mlp.gate_proj.weight": "
|
| 841 |
-
"layers.32.mlp.gate_proj.scales": "
|
| 842 |
-
"layers.32.mlp.gate_proj.biases": "
|
| 843 |
-
"layers.32.mlp.up_proj.weight": "
|
| 844 |
-
"layers.32.mlp.up_proj.scales": "
|
| 845 |
-
"layers.32.mlp.up_proj.biases": "
|
| 846 |
-
"layers.32.mlp.linear_fc2.weight": "
|
| 847 |
-
"layers.32.mlp.linear_fc2.scales": "
|
| 848 |
-
"layers.32.mlp.linear_fc2.biases": "
|
| 849 |
-
"layers.33.adaLN_sa_ln.weight": "
|
| 850 |
-
"layers.33.self_attention.to_q.weight": "
|
| 851 |
-
"layers.33.self_attention.to_q.scales": "
|
| 852 |
-
"layers.33.self_attention.to_q.biases": "
|
| 853 |
-
"layers.33.self_attention.to_k.weight": "
|
| 854 |
-
"layers.33.self_attention.to_k.scales": "
|
| 855 |
-
"layers.33.self_attention.to_k.biases": "
|
| 856 |
-
"layers.33.self_attention.to_v.weight": "
|
| 857 |
-
"layers.33.self_attention.to_v.scales": "
|
| 858 |
-
"layers.33.self_attention.to_v.biases": "
|
| 859 |
-
"layers.33.self_attention.norm_q.weight": "
|
| 860 |
-
"layers.33.self_attention.norm_k.weight": "
|
| 861 |
-
"layers.33.self_attention.to_out.0.weight": "
|
| 862 |
-
"layers.33.self_attention.to_out.0.scales": "
|
| 863 |
-
"layers.33.self_attention.to_out.0.biases": "
|
| 864 |
-
"layers.33.adaLN_mlp_ln.weight": "
|
| 865 |
-
"layers.33.mlp.gate_proj.weight": "
|
| 866 |
-
"layers.33.mlp.gate_proj.scales": "
|
| 867 |
-
"layers.33.mlp.gate_proj.biases": "
|
| 868 |
-
"layers.33.mlp.up_proj.weight": "
|
| 869 |
-
"layers.33.mlp.up_proj.scales": "
|
| 870 |
-
"layers.33.mlp.up_proj.biases": "
|
| 871 |
-
"layers.33.mlp.linear_fc2.weight": "
|
| 872 |
-
"layers.33.mlp.linear_fc2.scales": "
|
| 873 |
-
"layers.33.mlp.linear_fc2.biases": "
|
| 874 |
-
"layers.34.adaLN_sa_ln.weight": "
|
| 875 |
-
"layers.34.self_attention.to_q.weight": "
|
| 876 |
-
"layers.34.self_attention.to_q.scales": "
|
| 877 |
-
"layers.34.self_attention.to_q.biases": "
|
| 878 |
-
"layers.34.self_attention.to_k.weight": "
|
| 879 |
-
"layers.34.self_attention.to_k.scales": "
|
| 880 |
-
"layers.34.self_attention.to_k.biases": "
|
| 881 |
-
"layers.34.self_attention.to_v.weight": "
|
| 882 |
-
"layers.34.self_attention.to_v.scales": "
|
| 883 |
-
"layers.34.self_attention.to_v.biases": "
|
| 884 |
-
"layers.34.self_attention.norm_q.weight": "
|
| 885 |
-
"layers.34.self_attention.norm_k.weight": "
|
| 886 |
-
"layers.34.self_attention.to_out.0.weight": "
|
| 887 |
-
"layers.34.self_attention.to_out.0.scales": "
|
| 888 |
-
"layers.34.self_attention.to_out.0.biases": "
|
| 889 |
-
"layers.34.adaLN_mlp_ln.weight": "
|
| 890 |
-
"layers.34.mlp.gate_proj.weight": "
|
| 891 |
-
"layers.34.mlp.gate_proj.scales": "
|
| 892 |
-
"layers.34.mlp.gate_proj.biases": "
|
| 893 |
-
"layers.34.mlp.up_proj.weight": "
|
| 894 |
-
"layers.34.mlp.up_proj.scales": "
|
| 895 |
-
"layers.34.mlp.up_proj.biases": "
|
| 896 |
-
"layers.34.mlp.linear_fc2.weight": "
|
| 897 |
-
"layers.34.mlp.linear_fc2.scales": "
|
| 898 |
-
"layers.34.mlp.linear_fc2.biases": "
|
| 899 |
-
"layers.35.adaLN_sa_ln.weight": "
|
| 900 |
-
"layers.35.self_attention.to_q.weight": "
|
| 901 |
-
"layers.35.self_attention.to_q.scales": "
|
| 902 |
-
"layers.35.self_attention.to_q.biases": "
|
| 903 |
-
"layers.35.self_attention.to_k.weight": "
|
| 904 |
-
"layers.35.self_attention.to_k.scales": "
|
| 905 |
-
"layers.35.self_attention.to_k.biases": "
|
| 906 |
-
"layers.35.self_attention.to_v.weight": "
|
| 907 |
-
"layers.35.self_attention.to_v.scales": "
|
| 908 |
-
"layers.35.self_attention.to_v.biases": "
|
| 909 |
-
"layers.35.self_attention.norm_q.weight": "
|
| 910 |
-
"layers.35.self_attention.norm_k.weight": "
|
| 911 |
-
"layers.35.self_attention.to_out.0.weight": "
|
| 912 |
-
"layers.35.self_attention.to_out.0.scales": "
|
| 913 |
-
"layers.35.self_attention.to_out.0.biases": "
|
| 914 |
-
"layers.35.adaLN_mlp_ln.weight": "
|
| 915 |
-
"layers.35.mlp.gate_proj.weight": "
|
| 916 |
-
"layers.35.mlp.gate_proj.scales": "
|
| 917 |
-
"layers.35.mlp.gate_proj.biases": "
|
| 918 |
-
"layers.35.mlp.up_proj.weight": "
|
| 919 |
-
"layers.35.mlp.up_proj.scales": "
|
| 920 |
-
"layers.35.mlp.up_proj.biases": "
|
| 921 |
-
"layers.35.mlp.linear_fc2.weight": "
|
| 922 |
-
"layers.35.mlp.linear_fc2.scales": "
|
| 923 |
-
"layers.35.mlp.linear_fc2.biases": "
|
| 924 |
-
"final_norm.linear.weight": "
|
| 925 |
-
"final_norm.linear.scales": "
|
| 926 |
-
"final_norm.linear.biases": "
|
| 927 |
-
"final_norm.linear.bias": "
|
| 928 |
-
"final_linear.weight": "
|
| 929 |
-
"final_linear.scales": "
|
| 930 |
-
"final_linear.biases": "
|
| 931 |
-
"final_linear.bias": "
|
| 932 |
}
|
| 933 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"metadata": {
|
| 3 |
+
"quantization_level": "8",
|
| 4 |
"mflux_version": "0.18.5"
|
| 5 |
},
|
| 6 |
"weight_map": {
|
|
|
|
| 240 |
"layers.8.mlp.gate_proj.weight": "0.safetensors",
|
| 241 |
"layers.8.mlp.gate_proj.scales": "0.safetensors",
|
| 242 |
"layers.8.mlp.gate_proj.biases": "0.safetensors",
|
| 243 |
+
"layers.8.mlp.up_proj.weight": "1.safetensors",
|
| 244 |
+
"layers.8.mlp.up_proj.scales": "1.safetensors",
|
| 245 |
+
"layers.8.mlp.up_proj.biases": "1.safetensors",
|
| 246 |
+
"layers.8.mlp.linear_fc2.weight": "1.safetensors",
|
| 247 |
+
"layers.8.mlp.linear_fc2.scales": "1.safetensors",
|
| 248 |
+
"layers.8.mlp.linear_fc2.biases": "1.safetensors",
|
| 249 |
+
"layers.9.adaLN_sa_ln.weight": "1.safetensors",
|
| 250 |
+
"layers.9.self_attention.to_q.weight": "1.safetensors",
|
| 251 |
+
"layers.9.self_attention.to_q.scales": "1.safetensors",
|
| 252 |
+
"layers.9.self_attention.to_q.biases": "1.safetensors",
|
| 253 |
+
"layers.9.self_attention.to_k.weight": "1.safetensors",
|
| 254 |
+
"layers.9.self_attention.to_k.scales": "1.safetensors",
|
| 255 |
+
"layers.9.self_attention.to_k.biases": "1.safetensors",
|
| 256 |
+
"layers.9.self_attention.to_v.weight": "1.safetensors",
|
| 257 |
+
"layers.9.self_attention.to_v.scales": "1.safetensors",
|
| 258 |
+
"layers.9.self_attention.to_v.biases": "1.safetensors",
|
| 259 |
+
"layers.9.self_attention.norm_q.weight": "1.safetensors",
|
| 260 |
+
"layers.9.self_attention.norm_k.weight": "1.safetensors",
|
| 261 |
+
"layers.9.self_attention.to_out.0.weight": "1.safetensors",
|
| 262 |
+
"layers.9.self_attention.to_out.0.scales": "1.safetensors",
|
| 263 |
+
"layers.9.self_attention.to_out.0.biases": "1.safetensors",
|
| 264 |
+
"layers.9.adaLN_mlp_ln.weight": "1.safetensors",
|
| 265 |
+
"layers.9.mlp.gate_proj.weight": "1.safetensors",
|
| 266 |
+
"layers.9.mlp.gate_proj.scales": "1.safetensors",
|
| 267 |
+
"layers.9.mlp.gate_proj.biases": "1.safetensors",
|
| 268 |
+
"layers.9.mlp.up_proj.weight": "1.safetensors",
|
| 269 |
+
"layers.9.mlp.up_proj.scales": "1.safetensors",
|
| 270 |
+
"layers.9.mlp.up_proj.biases": "1.safetensors",
|
| 271 |
+
"layers.9.mlp.linear_fc2.weight": "1.safetensors",
|
| 272 |
+
"layers.9.mlp.linear_fc2.scales": "1.safetensors",
|
| 273 |
+
"layers.9.mlp.linear_fc2.biases": "1.safetensors",
|
| 274 |
+
"layers.10.adaLN_sa_ln.weight": "1.safetensors",
|
| 275 |
+
"layers.10.self_attention.to_q.weight": "1.safetensors",
|
| 276 |
+
"layers.10.self_attention.to_q.scales": "1.safetensors",
|
| 277 |
+
"layers.10.self_attention.to_q.biases": "1.safetensors",
|
| 278 |
+
"layers.10.self_attention.to_k.weight": "1.safetensors",
|
| 279 |
+
"layers.10.self_attention.to_k.scales": "1.safetensors",
|
| 280 |
+
"layers.10.self_attention.to_k.biases": "1.safetensors",
|
| 281 |
+
"layers.10.self_attention.to_v.weight": "1.safetensors",
|
| 282 |
+
"layers.10.self_attention.to_v.scales": "1.safetensors",
|
| 283 |
+
"layers.10.self_attention.to_v.biases": "1.safetensors",
|
| 284 |
+
"layers.10.self_attention.norm_q.weight": "1.safetensors",
|
| 285 |
+
"layers.10.self_attention.norm_k.weight": "1.safetensors",
|
| 286 |
+
"layers.10.self_attention.to_out.0.weight": "1.safetensors",
|
| 287 |
+
"layers.10.self_attention.to_out.0.scales": "1.safetensors",
|
| 288 |
+
"layers.10.self_attention.to_out.0.biases": "1.safetensors",
|
| 289 |
+
"layers.10.adaLN_mlp_ln.weight": "1.safetensors",
|
| 290 |
+
"layers.10.mlp.gate_proj.weight": "1.safetensors",
|
| 291 |
+
"layers.10.mlp.gate_proj.scales": "1.safetensors",
|
| 292 |
+
"layers.10.mlp.gate_proj.biases": "1.safetensors",
|
| 293 |
+
"layers.10.mlp.up_proj.weight": "1.safetensors",
|
| 294 |
+
"layers.10.mlp.up_proj.scales": "1.safetensors",
|
| 295 |
+
"layers.10.mlp.up_proj.biases": "1.safetensors",
|
| 296 |
+
"layers.10.mlp.linear_fc2.weight": "1.safetensors",
|
| 297 |
+
"layers.10.mlp.linear_fc2.scales": "1.safetensors",
|
| 298 |
+
"layers.10.mlp.linear_fc2.biases": "1.safetensors",
|
| 299 |
+
"layers.11.adaLN_sa_ln.weight": "1.safetensors",
|
| 300 |
+
"layers.11.self_attention.to_q.weight": "1.safetensors",
|
| 301 |
+
"layers.11.self_attention.to_q.scales": "1.safetensors",
|
| 302 |
+
"layers.11.self_attention.to_q.biases": "1.safetensors",
|
| 303 |
+
"layers.11.self_attention.to_k.weight": "1.safetensors",
|
| 304 |
+
"layers.11.self_attention.to_k.scales": "1.safetensors",
|
| 305 |
+
"layers.11.self_attention.to_k.biases": "1.safetensors",
|
| 306 |
+
"layers.11.self_attention.to_v.weight": "1.safetensors",
|
| 307 |
+
"layers.11.self_attention.to_v.scales": "1.safetensors",
|
| 308 |
+
"layers.11.self_attention.to_v.biases": "1.safetensors",
|
| 309 |
+
"layers.11.self_attention.norm_q.weight": "1.safetensors",
|
| 310 |
+
"layers.11.self_attention.norm_k.weight": "1.safetensors",
|
| 311 |
+
"layers.11.self_attention.to_out.0.weight": "1.safetensors",
|
| 312 |
+
"layers.11.self_attention.to_out.0.scales": "1.safetensors",
|
| 313 |
+
"layers.11.self_attention.to_out.0.biases": "1.safetensors",
|
| 314 |
+
"layers.11.adaLN_mlp_ln.weight": "1.safetensors",
|
| 315 |
+
"layers.11.mlp.gate_proj.weight": "1.safetensors",
|
| 316 |
+
"layers.11.mlp.gate_proj.scales": "1.safetensors",
|
| 317 |
+
"layers.11.mlp.gate_proj.biases": "1.safetensors",
|
| 318 |
+
"layers.11.mlp.up_proj.weight": "1.safetensors",
|
| 319 |
+
"layers.11.mlp.up_proj.scales": "1.safetensors",
|
| 320 |
+
"layers.11.mlp.up_proj.biases": "1.safetensors",
|
| 321 |
+
"layers.11.mlp.linear_fc2.weight": "1.safetensors",
|
| 322 |
+
"layers.11.mlp.linear_fc2.scales": "1.safetensors",
|
| 323 |
+
"layers.11.mlp.linear_fc2.biases": "1.safetensors",
|
| 324 |
+
"layers.12.adaLN_sa_ln.weight": "1.safetensors",
|
| 325 |
+
"layers.12.self_attention.to_q.weight": "1.safetensors",
|
| 326 |
+
"layers.12.self_attention.to_q.scales": "1.safetensors",
|
| 327 |
+
"layers.12.self_attention.to_q.biases": "1.safetensors",
|
| 328 |
+
"layers.12.self_attention.to_k.weight": "1.safetensors",
|
| 329 |
+
"layers.12.self_attention.to_k.scales": "1.safetensors",
|
| 330 |
+
"layers.12.self_attention.to_k.biases": "1.safetensors",
|
| 331 |
+
"layers.12.self_attention.to_v.weight": "1.safetensors",
|
| 332 |
+
"layers.12.self_attention.to_v.scales": "1.safetensors",
|
| 333 |
+
"layers.12.self_attention.to_v.biases": "1.safetensors",
|
| 334 |
+
"layers.12.self_attention.norm_q.weight": "1.safetensors",
|
| 335 |
+
"layers.12.self_attention.norm_k.weight": "1.safetensors",
|
| 336 |
+
"layers.12.self_attention.to_out.0.weight": "1.safetensors",
|
| 337 |
+
"layers.12.self_attention.to_out.0.scales": "1.safetensors",
|
| 338 |
+
"layers.12.self_attention.to_out.0.biases": "1.safetensors",
|
| 339 |
+
"layers.12.adaLN_mlp_ln.weight": "1.safetensors",
|
| 340 |
+
"layers.12.mlp.gate_proj.weight": "1.safetensors",
|
| 341 |
+
"layers.12.mlp.gate_proj.scales": "1.safetensors",
|
| 342 |
+
"layers.12.mlp.gate_proj.biases": "1.safetensors",
|
| 343 |
+
"layers.12.mlp.up_proj.weight": "1.safetensors",
|
| 344 |
+
"layers.12.mlp.up_proj.scales": "1.safetensors",
|
| 345 |
+
"layers.12.mlp.up_proj.biases": "1.safetensors",
|
| 346 |
+
"layers.12.mlp.linear_fc2.weight": "1.safetensors",
|
| 347 |
+
"layers.12.mlp.linear_fc2.scales": "1.safetensors",
|
| 348 |
+
"layers.12.mlp.linear_fc2.biases": "1.safetensors",
|
| 349 |
+
"layers.13.adaLN_sa_ln.weight": "1.safetensors",
|
| 350 |
+
"layers.13.self_attention.to_q.weight": "1.safetensors",
|
| 351 |
+
"layers.13.self_attention.to_q.scales": "1.safetensors",
|
| 352 |
+
"layers.13.self_attention.to_q.biases": "1.safetensors",
|
| 353 |
+
"layers.13.self_attention.to_k.weight": "1.safetensors",
|
| 354 |
+
"layers.13.self_attention.to_k.scales": "1.safetensors",
|
| 355 |
+
"layers.13.self_attention.to_k.biases": "1.safetensors",
|
| 356 |
+
"layers.13.self_attention.to_v.weight": "1.safetensors",
|
| 357 |
+
"layers.13.self_attention.to_v.scales": "1.safetensors",
|
| 358 |
+
"layers.13.self_attention.to_v.biases": "1.safetensors",
|
| 359 |
+
"layers.13.self_attention.norm_q.weight": "1.safetensors",
|
| 360 |
+
"layers.13.self_attention.norm_k.weight": "1.safetensors",
|
| 361 |
+
"layers.13.self_attention.to_out.0.weight": "1.safetensors",
|
| 362 |
+
"layers.13.self_attention.to_out.0.scales": "1.safetensors",
|
| 363 |
+
"layers.13.self_attention.to_out.0.biases": "1.safetensors",
|
| 364 |
+
"layers.13.adaLN_mlp_ln.weight": "1.safetensors",
|
| 365 |
+
"layers.13.mlp.gate_proj.weight": "1.safetensors",
|
| 366 |
+
"layers.13.mlp.gate_proj.scales": "1.safetensors",
|
| 367 |
+
"layers.13.mlp.gate_proj.biases": "1.safetensors",
|
| 368 |
+
"layers.13.mlp.up_proj.weight": "1.safetensors",
|
| 369 |
+
"layers.13.mlp.up_proj.scales": "1.safetensors",
|
| 370 |
+
"layers.13.mlp.up_proj.biases": "1.safetensors",
|
| 371 |
+
"layers.13.mlp.linear_fc2.weight": "1.safetensors",
|
| 372 |
+
"layers.13.mlp.linear_fc2.scales": "1.safetensors",
|
| 373 |
+
"layers.13.mlp.linear_fc2.biases": "1.safetensors",
|
| 374 |
+
"layers.14.adaLN_sa_ln.weight": "1.safetensors",
|
| 375 |
+
"layers.14.self_attention.to_q.weight": "1.safetensors",
|
| 376 |
+
"layers.14.self_attention.to_q.scales": "1.safetensors",
|
| 377 |
+
"layers.14.self_attention.to_q.biases": "1.safetensors",
|
| 378 |
+
"layers.14.self_attention.to_k.weight": "1.safetensors",
|
| 379 |
+
"layers.14.self_attention.to_k.scales": "1.safetensors",
|
| 380 |
+
"layers.14.self_attention.to_k.biases": "1.safetensors",
|
| 381 |
+
"layers.14.self_attention.to_v.weight": "1.safetensors",
|
| 382 |
+
"layers.14.self_attention.to_v.scales": "1.safetensors",
|
| 383 |
+
"layers.14.self_attention.to_v.biases": "1.safetensors",
|
| 384 |
+
"layers.14.self_attention.norm_q.weight": "1.safetensors",
|
| 385 |
+
"layers.14.self_attention.norm_k.weight": "1.safetensors",
|
| 386 |
"layers.14.self_attention.to_out.0.weight": "1.safetensors",
|
| 387 |
"layers.14.self_attention.to_out.0.scales": "1.safetensors",
|
| 388 |
"layers.14.self_attention.to_out.0.biases": "1.safetensors",
|
|
|
|
| 468 |
"layers.17.mlp.up_proj.weight": "1.safetensors",
|
| 469 |
"layers.17.mlp.up_proj.scales": "1.safetensors",
|
| 470 |
"layers.17.mlp.up_proj.biases": "1.safetensors",
|
| 471 |
+
"layers.17.mlp.linear_fc2.weight": "2.safetensors",
|
| 472 |
+
"layers.17.mlp.linear_fc2.scales": "2.safetensors",
|
| 473 |
+
"layers.17.mlp.linear_fc2.biases": "2.safetensors",
|
| 474 |
+
"layers.18.adaLN_sa_ln.weight": "2.safetensors",
|
| 475 |
+
"layers.18.self_attention.to_q.weight": "2.safetensors",
|
| 476 |
+
"layers.18.self_attention.to_q.scales": "2.safetensors",
|
| 477 |
+
"layers.18.self_attention.to_q.biases": "2.safetensors",
|
| 478 |
+
"layers.18.self_attention.to_k.weight": "2.safetensors",
|
| 479 |
+
"layers.18.self_attention.to_k.scales": "2.safetensors",
|
| 480 |
+
"layers.18.self_attention.to_k.biases": "2.safetensors",
|
| 481 |
+
"layers.18.self_attention.to_v.weight": "2.safetensors",
|
| 482 |
+
"layers.18.self_attention.to_v.scales": "2.safetensors",
|
| 483 |
+
"layers.18.self_attention.to_v.biases": "2.safetensors",
|
| 484 |
+
"layers.18.self_attention.norm_q.weight": "2.safetensors",
|
| 485 |
+
"layers.18.self_attention.norm_k.weight": "2.safetensors",
|
| 486 |
+
"layers.18.self_attention.to_out.0.weight": "2.safetensors",
|
| 487 |
+
"layers.18.self_attention.to_out.0.scales": "2.safetensors",
|
| 488 |
+
"layers.18.self_attention.to_out.0.biases": "2.safetensors",
|
| 489 |
+
"layers.18.adaLN_mlp_ln.weight": "2.safetensors",
|
| 490 |
+
"layers.18.mlp.gate_proj.weight": "2.safetensors",
|
| 491 |
+
"layers.18.mlp.gate_proj.scales": "2.safetensors",
|
| 492 |
+
"layers.18.mlp.gate_proj.biases": "2.safetensors",
|
| 493 |
+
"layers.18.mlp.up_proj.weight": "2.safetensors",
|
| 494 |
+
"layers.18.mlp.up_proj.scales": "2.safetensors",
|
| 495 |
+
"layers.18.mlp.up_proj.biases": "2.safetensors",
|
| 496 |
+
"layers.18.mlp.linear_fc2.weight": "2.safetensors",
|
| 497 |
+
"layers.18.mlp.linear_fc2.scales": "2.safetensors",
|
| 498 |
+
"layers.18.mlp.linear_fc2.biases": "2.safetensors",
|
| 499 |
+
"layers.19.adaLN_sa_ln.weight": "2.safetensors",
|
| 500 |
+
"layers.19.self_attention.to_q.weight": "2.safetensors",
|
| 501 |
+
"layers.19.self_attention.to_q.scales": "2.safetensors",
|
| 502 |
+
"layers.19.self_attention.to_q.biases": "2.safetensors",
|
| 503 |
+
"layers.19.self_attention.to_k.weight": "2.safetensors",
|
| 504 |
+
"layers.19.self_attention.to_k.scales": "2.safetensors",
|
| 505 |
+
"layers.19.self_attention.to_k.biases": "2.safetensors",
|
| 506 |
+
"layers.19.self_attention.to_v.weight": "2.safetensors",
|
| 507 |
+
"layers.19.self_attention.to_v.scales": "2.safetensors",
|
| 508 |
+
"layers.19.self_attention.to_v.biases": "2.safetensors",
|
| 509 |
+
"layers.19.self_attention.norm_q.weight": "2.safetensors",
|
| 510 |
+
"layers.19.self_attention.norm_k.weight": "2.safetensors",
|
| 511 |
+
"layers.19.self_attention.to_out.0.weight": "2.safetensors",
|
| 512 |
+
"layers.19.self_attention.to_out.0.scales": "2.safetensors",
|
| 513 |
+
"layers.19.self_attention.to_out.0.biases": "2.safetensors",
|
| 514 |
+
"layers.19.adaLN_mlp_ln.weight": "2.safetensors",
|
| 515 |
+
"layers.19.mlp.gate_proj.weight": "2.safetensors",
|
| 516 |
+
"layers.19.mlp.gate_proj.scales": "2.safetensors",
|
| 517 |
+
"layers.19.mlp.gate_proj.biases": "2.safetensors",
|
| 518 |
+
"layers.19.mlp.up_proj.weight": "2.safetensors",
|
| 519 |
+
"layers.19.mlp.up_proj.scales": "2.safetensors",
|
| 520 |
+
"layers.19.mlp.up_proj.biases": "2.safetensors",
|
| 521 |
+
"layers.19.mlp.linear_fc2.weight": "2.safetensors",
|
| 522 |
+
"layers.19.mlp.linear_fc2.scales": "2.safetensors",
|
| 523 |
+
"layers.19.mlp.linear_fc2.biases": "2.safetensors",
|
| 524 |
+
"layers.20.adaLN_sa_ln.weight": "2.safetensors",
|
| 525 |
+
"layers.20.self_attention.to_q.weight": "2.safetensors",
|
| 526 |
+
"layers.20.self_attention.to_q.scales": "2.safetensors",
|
| 527 |
+
"layers.20.self_attention.to_q.biases": "2.safetensors",
|
| 528 |
+
"layers.20.self_attention.to_k.weight": "2.safetensors",
|
| 529 |
+
"layers.20.self_attention.to_k.scales": "2.safetensors",
|
| 530 |
+
"layers.20.self_attention.to_k.biases": "2.safetensors",
|
| 531 |
+
"layers.20.self_attention.to_v.weight": "2.safetensors",
|
| 532 |
+
"layers.20.self_attention.to_v.scales": "2.safetensors",
|
| 533 |
+
"layers.20.self_attention.to_v.biases": "2.safetensors",
|
| 534 |
+
"layers.20.self_attention.norm_q.weight": "2.safetensors",
|
| 535 |
+
"layers.20.self_attention.norm_k.weight": "2.safetensors",
|
| 536 |
+
"layers.20.self_attention.to_out.0.weight": "2.safetensors",
|
| 537 |
+
"layers.20.self_attention.to_out.0.scales": "2.safetensors",
|
| 538 |
+
"layers.20.self_attention.to_out.0.biases": "2.safetensors",
|
| 539 |
+
"layers.20.adaLN_mlp_ln.weight": "2.safetensors",
|
| 540 |
+
"layers.20.mlp.gate_proj.weight": "2.safetensors",
|
| 541 |
+
"layers.20.mlp.gate_proj.scales": "2.safetensors",
|
| 542 |
+
"layers.20.mlp.gate_proj.biases": "2.safetensors",
|
| 543 |
+
"layers.20.mlp.up_proj.weight": "2.safetensors",
|
| 544 |
+
"layers.20.mlp.up_proj.scales": "2.safetensors",
|
| 545 |
+
"layers.20.mlp.up_proj.biases": "2.safetensors",
|
| 546 |
+
"layers.20.mlp.linear_fc2.weight": "2.safetensors",
|
| 547 |
+
"layers.20.mlp.linear_fc2.scales": "2.safetensors",
|
| 548 |
+
"layers.20.mlp.linear_fc2.biases": "2.safetensors",
|
| 549 |
+
"layers.21.adaLN_sa_ln.weight": "2.safetensors",
|
| 550 |
+
"layers.21.self_attention.to_q.weight": "2.safetensors",
|
| 551 |
+
"layers.21.self_attention.to_q.scales": "2.safetensors",
|
| 552 |
+
"layers.21.self_attention.to_q.biases": "2.safetensors",
|
| 553 |
+
"layers.21.self_attention.to_k.weight": "2.safetensors",
|
| 554 |
+
"layers.21.self_attention.to_k.scales": "2.safetensors",
|
| 555 |
+
"layers.21.self_attention.to_k.biases": "2.safetensors",
|
| 556 |
+
"layers.21.self_attention.to_v.weight": "2.safetensors",
|
| 557 |
+
"layers.21.self_attention.to_v.scales": "2.safetensors",
|
| 558 |
+
"layers.21.self_attention.to_v.biases": "2.safetensors",
|
| 559 |
+
"layers.21.self_attention.norm_q.weight": "2.safetensors",
|
| 560 |
+
"layers.21.self_attention.norm_k.weight": "2.safetensors",
|
| 561 |
+
"layers.21.self_attention.to_out.0.weight": "2.safetensors",
|
| 562 |
+
"layers.21.self_attention.to_out.0.scales": "2.safetensors",
|
| 563 |
+
"layers.21.self_attention.to_out.0.biases": "2.safetensors",
|
| 564 |
+
"layers.21.adaLN_mlp_ln.weight": "2.safetensors",
|
| 565 |
+
"layers.21.mlp.gate_proj.weight": "2.safetensors",
|
| 566 |
+
"layers.21.mlp.gate_proj.scales": "2.safetensors",
|
| 567 |
+
"layers.21.mlp.gate_proj.biases": "2.safetensors",
|
| 568 |
+
"layers.21.mlp.up_proj.weight": "2.safetensors",
|
| 569 |
+
"layers.21.mlp.up_proj.scales": "2.safetensors",
|
| 570 |
+
"layers.21.mlp.up_proj.biases": "2.safetensors",
|
| 571 |
+
"layers.21.mlp.linear_fc2.weight": "2.safetensors",
|
| 572 |
+
"layers.21.mlp.linear_fc2.scales": "2.safetensors",
|
| 573 |
+
"layers.21.mlp.linear_fc2.biases": "2.safetensors",
|
| 574 |
+
"layers.22.adaLN_sa_ln.weight": "2.safetensors",
|
| 575 |
+
"layers.22.self_attention.to_q.weight": "2.safetensors",
|
| 576 |
+
"layers.22.self_attention.to_q.scales": "2.safetensors",
|
| 577 |
+
"layers.22.self_attention.to_q.biases": "2.safetensors",
|
| 578 |
+
"layers.22.self_attention.to_k.weight": "2.safetensors",
|
| 579 |
+
"layers.22.self_attention.to_k.scales": "2.safetensors",
|
| 580 |
+
"layers.22.self_attention.to_k.biases": "2.safetensors",
|
| 581 |
+
"layers.22.self_attention.to_v.weight": "2.safetensors",
|
| 582 |
+
"layers.22.self_attention.to_v.scales": "2.safetensors",
|
| 583 |
+
"layers.22.self_attention.to_v.biases": "2.safetensors",
|
| 584 |
+
"layers.22.self_attention.norm_q.weight": "2.safetensors",
|
| 585 |
+
"layers.22.self_attention.norm_k.weight": "2.safetensors",
|
| 586 |
+
"layers.22.self_attention.to_out.0.weight": "2.safetensors",
|
| 587 |
+
"layers.22.self_attention.to_out.0.scales": "2.safetensors",
|
| 588 |
+
"layers.22.self_attention.to_out.0.biases": "2.safetensors",
|
| 589 |
+
"layers.22.adaLN_mlp_ln.weight": "2.safetensors",
|
| 590 |
+
"layers.22.mlp.gate_proj.weight": "2.safetensors",
|
| 591 |
+
"layers.22.mlp.gate_proj.scales": "2.safetensors",
|
| 592 |
+
"layers.22.mlp.gate_proj.biases": "2.safetensors",
|
| 593 |
+
"layers.22.mlp.up_proj.weight": "2.safetensors",
|
| 594 |
+
"layers.22.mlp.up_proj.scales": "2.safetensors",
|
| 595 |
+
"layers.22.mlp.up_proj.biases": "2.safetensors",
|
| 596 |
+
"layers.22.mlp.linear_fc2.weight": "2.safetensors",
|
| 597 |
+
"layers.22.mlp.linear_fc2.scales": "2.safetensors",
|
| 598 |
+
"layers.22.mlp.linear_fc2.biases": "2.safetensors",
|
| 599 |
+
"layers.23.adaLN_sa_ln.weight": "2.safetensors",
|
| 600 |
+
"layers.23.self_attention.to_q.weight": "2.safetensors",
|
| 601 |
+
"layers.23.self_attention.to_q.scales": "2.safetensors",
|
| 602 |
+
"layers.23.self_attention.to_q.biases": "2.safetensors",
|
| 603 |
+
"layers.23.self_attention.to_k.weight": "2.safetensors",
|
| 604 |
+
"layers.23.self_attention.to_k.scales": "2.safetensors",
|
| 605 |
+
"layers.23.self_attention.to_k.biases": "2.safetensors",
|
| 606 |
+
"layers.23.self_attention.to_v.weight": "2.safetensors",
|
| 607 |
+
"layers.23.self_attention.to_v.scales": "2.safetensors",
|
| 608 |
+
"layers.23.self_attention.to_v.biases": "2.safetensors",
|
| 609 |
+
"layers.23.self_attention.norm_q.weight": "2.safetensors",
|
| 610 |
+
"layers.23.self_attention.norm_k.weight": "2.safetensors",
|
| 611 |
+
"layers.23.self_attention.to_out.0.weight": "2.safetensors",
|
| 612 |
+
"layers.23.self_attention.to_out.0.scales": "2.safetensors",
|
| 613 |
+
"layers.23.self_attention.to_out.0.biases": "2.safetensors",
|
| 614 |
+
"layers.23.adaLN_mlp_ln.weight": "2.safetensors",
|
| 615 |
+
"layers.23.mlp.gate_proj.weight": "2.safetensors",
|
| 616 |
+
"layers.23.mlp.gate_proj.scales": "2.safetensors",
|
| 617 |
+
"layers.23.mlp.gate_proj.biases": "2.safetensors",
|
| 618 |
+
"layers.23.mlp.up_proj.weight": "2.safetensors",
|
| 619 |
+
"layers.23.mlp.up_proj.scales": "2.safetensors",
|
| 620 |
+
"layers.23.mlp.up_proj.biases": "2.safetensors",
|
| 621 |
+
"layers.23.mlp.linear_fc2.weight": "2.safetensors",
|
| 622 |
+
"layers.23.mlp.linear_fc2.scales": "2.safetensors",
|
| 623 |
+
"layers.23.mlp.linear_fc2.biases": "2.safetensors",
|
| 624 |
+
"layers.24.adaLN_sa_ln.weight": "2.safetensors",
|
| 625 |
+
"layers.24.self_attention.to_q.weight": "2.safetensors",
|
| 626 |
+
"layers.24.self_attention.to_q.scales": "2.safetensors",
|
| 627 |
+
"layers.24.self_attention.to_q.biases": "2.safetensors",
|
| 628 |
+
"layers.24.self_attention.to_k.weight": "2.safetensors",
|
| 629 |
+
"layers.24.self_attention.to_k.scales": "2.safetensors",
|
| 630 |
+
"layers.24.self_attention.to_k.biases": "2.safetensors",
|
| 631 |
+
"layers.24.self_attention.to_v.weight": "2.safetensors",
|
| 632 |
+
"layers.24.self_attention.to_v.scales": "2.safetensors",
|
| 633 |
+
"layers.24.self_attention.to_v.biases": "2.safetensors",
|
| 634 |
+
"layers.24.self_attention.norm_q.weight": "2.safetensors",
|
| 635 |
+
"layers.24.self_attention.norm_k.weight": "2.safetensors",
|
| 636 |
+
"layers.24.self_attention.to_out.0.weight": "2.safetensors",
|
| 637 |
+
"layers.24.self_attention.to_out.0.scales": "2.safetensors",
|
| 638 |
+
"layers.24.self_attention.to_out.0.biases": "2.safetensors",
|
| 639 |
+
"layers.24.adaLN_mlp_ln.weight": "2.safetensors",
|
| 640 |
+
"layers.24.mlp.gate_proj.weight": "2.safetensors",
|
| 641 |
+
"layers.24.mlp.gate_proj.scales": "2.safetensors",
|
| 642 |
+
"layers.24.mlp.gate_proj.biases": "2.safetensors",
|
| 643 |
+
"layers.24.mlp.up_proj.weight": "2.safetensors",
|
| 644 |
+
"layers.24.mlp.up_proj.scales": "2.safetensors",
|
| 645 |
+
"layers.24.mlp.up_proj.biases": "2.safetensors",
|
| 646 |
+
"layers.24.mlp.linear_fc2.weight": "2.safetensors",
|
| 647 |
+
"layers.24.mlp.linear_fc2.scales": "2.safetensors",
|
| 648 |
+
"layers.24.mlp.linear_fc2.biases": "2.safetensors",
|
| 649 |
+
"layers.25.adaLN_sa_ln.weight": "2.safetensors",
|
| 650 |
+
"layers.25.self_attention.to_q.weight": "2.safetensors",
|
| 651 |
+
"layers.25.self_attention.to_q.scales": "2.safetensors",
|
| 652 |
+
"layers.25.self_attention.to_q.biases": "2.safetensors",
|
| 653 |
+
"layers.25.self_attention.to_k.weight": "2.safetensors",
|
| 654 |
+
"layers.25.self_attention.to_k.scales": "2.safetensors",
|
| 655 |
+
"layers.25.self_attention.to_k.biases": "2.safetensors",
|
| 656 |
+
"layers.25.self_attention.to_v.weight": "2.safetensors",
|
| 657 |
+
"layers.25.self_attention.to_v.scales": "2.safetensors",
|
| 658 |
+
"layers.25.self_attention.to_v.biases": "2.safetensors",
|
| 659 |
+
"layers.25.self_attention.norm_q.weight": "2.safetensors",
|
| 660 |
+
"layers.25.self_attention.norm_k.weight": "2.safetensors",
|
| 661 |
+
"layers.25.self_attention.to_out.0.weight": "2.safetensors",
|
| 662 |
+
"layers.25.self_attention.to_out.0.scales": "2.safetensors",
|
| 663 |
+
"layers.25.self_attention.to_out.0.biases": "2.safetensors",
|
| 664 |
+
"layers.25.adaLN_mlp_ln.weight": "2.safetensors",
|
| 665 |
+
"layers.25.mlp.gate_proj.weight": "2.safetensors",
|
| 666 |
+
"layers.25.mlp.gate_proj.scales": "2.safetensors",
|
| 667 |
+
"layers.25.mlp.gate_proj.biases": "2.safetensors",
|
| 668 |
+
"layers.25.mlp.up_proj.weight": "2.safetensors",
|
| 669 |
+
"layers.25.mlp.up_proj.scales": "2.safetensors",
|
| 670 |
+
"layers.25.mlp.up_proj.biases": "2.safetensors",
|
| 671 |
+
"layers.25.mlp.linear_fc2.weight": "2.safetensors",
|
| 672 |
+
"layers.25.mlp.linear_fc2.scales": "2.safetensors",
|
| 673 |
+
"layers.25.mlp.linear_fc2.biases": "2.safetensors",
|
| 674 |
+
"layers.26.adaLN_sa_ln.weight": "2.safetensors",
|
| 675 |
+
"layers.26.self_attention.to_q.weight": "2.safetensors",
|
| 676 |
+
"layers.26.self_attention.to_q.scales": "2.safetensors",
|
| 677 |
+
"layers.26.self_attention.to_q.biases": "2.safetensors",
|
| 678 |
+
"layers.26.self_attention.to_k.weight": "2.safetensors",
|
| 679 |
+
"layers.26.self_attention.to_k.scales": "2.safetensors",
|
| 680 |
+
"layers.26.self_attention.to_k.biases": "2.safetensors",
|
| 681 |
+
"layers.26.self_attention.to_v.weight": "2.safetensors",
|
| 682 |
+
"layers.26.self_attention.to_v.scales": "2.safetensors",
|
| 683 |
+
"layers.26.self_attention.to_v.biases": "2.safetensors",
|
| 684 |
+
"layers.26.self_attention.norm_q.weight": "2.safetensors",
|
| 685 |
+
"layers.26.self_attention.norm_k.weight": "2.safetensors",
|
| 686 |
+
"layers.26.self_attention.to_out.0.weight": "2.safetensors",
|
| 687 |
+
"layers.26.self_attention.to_out.0.scales": "2.safetensors",
|
| 688 |
+
"layers.26.self_attention.to_out.0.biases": "2.safetensors",
|
| 689 |
+
"layers.26.adaLN_mlp_ln.weight": "2.safetensors",
|
| 690 |
+
"layers.26.mlp.gate_proj.weight": "2.safetensors",
|
| 691 |
+
"layers.26.mlp.gate_proj.scales": "2.safetensors",
|
| 692 |
+
"layers.26.mlp.gate_proj.biases": "2.safetensors",
|
| 693 |
+
"layers.26.mlp.up_proj.weight": "2.safetensors",
|
| 694 |
+
"layers.26.mlp.up_proj.scales": "2.safetensors",
|
| 695 |
+
"layers.26.mlp.up_proj.biases": "2.safetensors",
|
| 696 |
+
"layers.26.mlp.linear_fc2.weight": "2.safetensors",
|
| 697 |
+
"layers.26.mlp.linear_fc2.scales": "2.safetensors",
|
| 698 |
+
"layers.26.mlp.linear_fc2.biases": "2.safetensors",
|
| 699 |
+
"layers.27.adaLN_sa_ln.weight": "2.safetensors",
|
| 700 |
+
"layers.27.self_attention.to_q.weight": "3.safetensors",
|
| 701 |
+
"layers.27.self_attention.to_q.scales": "3.safetensors",
|
| 702 |
+
"layers.27.self_attention.to_q.biases": "3.safetensors",
|
| 703 |
+
"layers.27.self_attention.to_k.weight": "3.safetensors",
|
| 704 |
+
"layers.27.self_attention.to_k.scales": "3.safetensors",
|
| 705 |
+
"layers.27.self_attention.to_k.biases": "3.safetensors",
|
| 706 |
+
"layers.27.self_attention.to_v.weight": "3.safetensors",
|
| 707 |
+
"layers.27.self_attention.to_v.scales": "3.safetensors",
|
| 708 |
+
"layers.27.self_attention.to_v.biases": "3.safetensors",
|
| 709 |
+
"layers.27.self_attention.norm_q.weight": "3.safetensors",
|
| 710 |
+
"layers.27.self_attention.norm_k.weight": "3.safetensors",
|
| 711 |
+
"layers.27.self_attention.to_out.0.weight": "3.safetensors",
|
| 712 |
+
"layers.27.self_attention.to_out.0.scales": "3.safetensors",
|
| 713 |
+
"layers.27.self_attention.to_out.0.biases": "3.safetensors",
|
| 714 |
+
"layers.27.adaLN_mlp_ln.weight": "3.safetensors",
|
| 715 |
+
"layers.27.mlp.gate_proj.weight": "3.safetensors",
|
| 716 |
+
"layers.27.mlp.gate_proj.scales": "3.safetensors",
|
| 717 |
+
"layers.27.mlp.gate_proj.biases": "3.safetensors",
|
| 718 |
+
"layers.27.mlp.up_proj.weight": "3.safetensors",
|
| 719 |
+
"layers.27.mlp.up_proj.scales": "3.safetensors",
|
| 720 |
+
"layers.27.mlp.up_proj.biases": "3.safetensors",
|
| 721 |
+
"layers.27.mlp.linear_fc2.weight": "3.safetensors",
|
| 722 |
+
"layers.27.mlp.linear_fc2.scales": "3.safetensors",
|
| 723 |
+
"layers.27.mlp.linear_fc2.biases": "3.safetensors",
|
| 724 |
+
"layers.28.adaLN_sa_ln.weight": "3.safetensors",
|
| 725 |
+
"layers.28.self_attention.to_q.weight": "3.safetensors",
|
| 726 |
+
"layers.28.self_attention.to_q.scales": "3.safetensors",
|
| 727 |
+
"layers.28.self_attention.to_q.biases": "3.safetensors",
|
| 728 |
+
"layers.28.self_attention.to_k.weight": "3.safetensors",
|
| 729 |
+
"layers.28.self_attention.to_k.scales": "3.safetensors",
|
| 730 |
+
"layers.28.self_attention.to_k.biases": "3.safetensors",
|
| 731 |
+
"layers.28.self_attention.to_v.weight": "3.safetensors",
|
| 732 |
+
"layers.28.self_attention.to_v.scales": "3.safetensors",
|
| 733 |
+
"layers.28.self_attention.to_v.biases": "3.safetensors",
|
| 734 |
+
"layers.28.self_attention.norm_q.weight": "3.safetensors",
|
| 735 |
+
"layers.28.self_attention.norm_k.weight": "3.safetensors",
|
| 736 |
+
"layers.28.self_attention.to_out.0.weight": "3.safetensors",
|
| 737 |
+
"layers.28.self_attention.to_out.0.scales": "3.safetensors",
|
| 738 |
+
"layers.28.self_attention.to_out.0.biases": "3.safetensors",
|
| 739 |
+
"layers.28.adaLN_mlp_ln.weight": "3.safetensors",
|
| 740 |
+
"layers.28.mlp.gate_proj.weight": "3.safetensors",
|
| 741 |
+
"layers.28.mlp.gate_proj.scales": "3.safetensors",
|
| 742 |
+
"layers.28.mlp.gate_proj.biases": "3.safetensors",
|
| 743 |
+
"layers.28.mlp.up_proj.weight": "3.safetensors",
|
| 744 |
+
"layers.28.mlp.up_proj.scales": "3.safetensors",
|
| 745 |
+
"layers.28.mlp.up_proj.biases": "3.safetensors",
|
| 746 |
+
"layers.28.mlp.linear_fc2.weight": "3.safetensors",
|
| 747 |
+
"layers.28.mlp.linear_fc2.scales": "3.safetensors",
|
| 748 |
+
"layers.28.mlp.linear_fc2.biases": "3.safetensors",
|
| 749 |
+
"layers.29.adaLN_sa_ln.weight": "3.safetensors",
|
| 750 |
+
"layers.29.self_attention.to_q.weight": "3.safetensors",
|
| 751 |
+
"layers.29.self_attention.to_q.scales": "3.safetensors",
|
| 752 |
+
"layers.29.self_attention.to_q.biases": "3.safetensors",
|
| 753 |
+
"layers.29.self_attention.to_k.weight": "3.safetensors",
|
| 754 |
+
"layers.29.self_attention.to_k.scales": "3.safetensors",
|
| 755 |
+
"layers.29.self_attention.to_k.biases": "3.safetensors",
|
| 756 |
+
"layers.29.self_attention.to_v.weight": "3.safetensors",
|
| 757 |
+
"layers.29.self_attention.to_v.scales": "3.safetensors",
|
| 758 |
+
"layers.29.self_attention.to_v.biases": "3.safetensors",
|
| 759 |
+
"layers.29.self_attention.norm_q.weight": "3.safetensors",
|
| 760 |
+
"layers.29.self_attention.norm_k.weight": "3.safetensors",
|
| 761 |
+
"layers.29.self_attention.to_out.0.weight": "3.safetensors",
|
| 762 |
+
"layers.29.self_attention.to_out.0.scales": "3.safetensors",
|
| 763 |
+
"layers.29.self_attention.to_out.0.biases": "3.safetensors",
|
| 764 |
+
"layers.29.adaLN_mlp_ln.weight": "3.safetensors",
|
| 765 |
+
"layers.29.mlp.gate_proj.weight": "3.safetensors",
|
| 766 |
+
"layers.29.mlp.gate_proj.scales": "3.safetensors",
|
| 767 |
+
"layers.29.mlp.gate_proj.biases": "3.safetensors",
|
| 768 |
+
"layers.29.mlp.up_proj.weight": "3.safetensors",
|
| 769 |
+
"layers.29.mlp.up_proj.scales": "3.safetensors",
|
| 770 |
+
"layers.29.mlp.up_proj.biases": "3.safetensors",
|
| 771 |
+
"layers.29.mlp.linear_fc2.weight": "3.safetensors",
|
| 772 |
+
"layers.29.mlp.linear_fc2.scales": "3.safetensors",
|
| 773 |
+
"layers.29.mlp.linear_fc2.biases": "3.safetensors",
|
| 774 |
+
"layers.30.adaLN_sa_ln.weight": "3.safetensors",
|
| 775 |
+
"layers.30.self_attention.to_q.weight": "3.safetensors",
|
| 776 |
+
"layers.30.self_attention.to_q.scales": "3.safetensors",
|
| 777 |
+
"layers.30.self_attention.to_q.biases": "3.safetensors",
|
| 778 |
+
"layers.30.self_attention.to_k.weight": "3.safetensors",
|
| 779 |
+
"layers.30.self_attention.to_k.scales": "3.safetensors",
|
| 780 |
+
"layers.30.self_attention.to_k.biases": "3.safetensors",
|
| 781 |
+
"layers.30.self_attention.to_v.weight": "3.safetensors",
|
| 782 |
+
"layers.30.self_attention.to_v.scales": "3.safetensors",
|
| 783 |
+
"layers.30.self_attention.to_v.biases": "3.safetensors",
|
| 784 |
+
"layers.30.self_attention.norm_q.weight": "3.safetensors",
|
| 785 |
+
"layers.30.self_attention.norm_k.weight": "3.safetensors",
|
| 786 |
+
"layers.30.self_attention.to_out.0.weight": "3.safetensors",
|
| 787 |
+
"layers.30.self_attention.to_out.0.scales": "3.safetensors",
|
| 788 |
+
"layers.30.self_attention.to_out.0.biases": "3.safetensors",
|
| 789 |
+
"layers.30.adaLN_mlp_ln.weight": "3.safetensors",
|
| 790 |
+
"layers.30.mlp.gate_proj.weight": "3.safetensors",
|
| 791 |
+
"layers.30.mlp.gate_proj.scales": "3.safetensors",
|
| 792 |
+
"layers.30.mlp.gate_proj.biases": "3.safetensors",
|
| 793 |
+
"layers.30.mlp.up_proj.weight": "3.safetensors",
|
| 794 |
+
"layers.30.mlp.up_proj.scales": "3.safetensors",
|
| 795 |
+
"layers.30.mlp.up_proj.biases": "3.safetensors",
|
| 796 |
+
"layers.30.mlp.linear_fc2.weight": "3.safetensors",
|
| 797 |
+
"layers.30.mlp.linear_fc2.scales": "3.safetensors",
|
| 798 |
+
"layers.30.mlp.linear_fc2.biases": "3.safetensors",
|
| 799 |
+
"layers.31.adaLN_sa_ln.weight": "3.safetensors",
|
| 800 |
+
"layers.31.self_attention.to_q.weight": "3.safetensors",
|
| 801 |
+
"layers.31.self_attention.to_q.scales": "3.safetensors",
|
| 802 |
+
"layers.31.self_attention.to_q.biases": "3.safetensors",
|
| 803 |
+
"layers.31.self_attention.to_k.weight": "3.safetensors",
|
| 804 |
+
"layers.31.self_attention.to_k.scales": "3.safetensors",
|
| 805 |
+
"layers.31.self_attention.to_k.biases": "3.safetensors",
|
| 806 |
+
"layers.31.self_attention.to_v.weight": "3.safetensors",
|
| 807 |
+
"layers.31.self_attention.to_v.scales": "3.safetensors",
|
| 808 |
+
"layers.31.self_attention.to_v.biases": "3.safetensors",
|
| 809 |
+
"layers.31.self_attention.norm_q.weight": "3.safetensors",
|
| 810 |
+
"layers.31.self_attention.norm_k.weight": "3.safetensors",
|
| 811 |
+
"layers.31.self_attention.to_out.0.weight": "3.safetensors",
|
| 812 |
+
"layers.31.self_attention.to_out.0.scales": "3.safetensors",
|
| 813 |
+
"layers.31.self_attention.to_out.0.biases": "3.safetensors",
|
| 814 |
+
"layers.31.adaLN_mlp_ln.weight": "3.safetensors",
|
| 815 |
+
"layers.31.mlp.gate_proj.weight": "3.safetensors",
|
| 816 |
+
"layers.31.mlp.gate_proj.scales": "3.safetensors",
|
| 817 |
+
"layers.31.mlp.gate_proj.biases": "3.safetensors",
|
| 818 |
+
"layers.31.mlp.up_proj.weight": "3.safetensors",
|
| 819 |
+
"layers.31.mlp.up_proj.scales": "3.safetensors",
|
| 820 |
+
"layers.31.mlp.up_proj.biases": "3.safetensors",
|
| 821 |
+
"layers.31.mlp.linear_fc2.weight": "3.safetensors",
|
| 822 |
+
"layers.31.mlp.linear_fc2.scales": "3.safetensors",
|
| 823 |
+
"layers.31.mlp.linear_fc2.biases": "3.safetensors",
|
| 824 |
+
"layers.32.adaLN_sa_ln.weight": "3.safetensors",
|
| 825 |
+
"layers.32.self_attention.to_q.weight": "3.safetensors",
|
| 826 |
+
"layers.32.self_attention.to_q.scales": "3.safetensors",
|
| 827 |
+
"layers.32.self_attention.to_q.biases": "3.safetensors",
|
| 828 |
+
"layers.32.self_attention.to_k.weight": "3.safetensors",
|
| 829 |
+
"layers.32.self_attention.to_k.scales": "3.safetensors",
|
| 830 |
+
"layers.32.self_attention.to_k.biases": "3.safetensors",
|
| 831 |
+
"layers.32.self_attention.to_v.weight": "3.safetensors",
|
| 832 |
+
"layers.32.self_attention.to_v.scales": "3.safetensors",
|
| 833 |
+
"layers.32.self_attention.to_v.biases": "3.safetensors",
|
| 834 |
+
"layers.32.self_attention.norm_q.weight": "3.safetensors",
|
| 835 |
+
"layers.32.self_attention.norm_k.weight": "3.safetensors",
|
| 836 |
+
"layers.32.self_attention.to_out.0.weight": "3.safetensors",
|
| 837 |
+
"layers.32.self_attention.to_out.0.scales": "3.safetensors",
|
| 838 |
+
"layers.32.self_attention.to_out.0.biases": "3.safetensors",
|
| 839 |
+
"layers.32.adaLN_mlp_ln.weight": "3.safetensors",
|
| 840 |
+
"layers.32.mlp.gate_proj.weight": "3.safetensors",
|
| 841 |
+
"layers.32.mlp.gate_proj.scales": "3.safetensors",
|
| 842 |
+
"layers.32.mlp.gate_proj.biases": "3.safetensors",
|
| 843 |
+
"layers.32.mlp.up_proj.weight": "3.safetensors",
|
| 844 |
+
"layers.32.mlp.up_proj.scales": "3.safetensors",
|
| 845 |
+
"layers.32.mlp.up_proj.biases": "3.safetensors",
|
| 846 |
+
"layers.32.mlp.linear_fc2.weight": "3.safetensors",
|
| 847 |
+
"layers.32.mlp.linear_fc2.scales": "3.safetensors",
|
| 848 |
+
"layers.32.mlp.linear_fc2.biases": "3.safetensors",
|
| 849 |
+
"layers.33.adaLN_sa_ln.weight": "3.safetensors",
|
| 850 |
+
"layers.33.self_attention.to_q.weight": "3.safetensors",
|
| 851 |
+
"layers.33.self_attention.to_q.scales": "3.safetensors",
|
| 852 |
+
"layers.33.self_attention.to_q.biases": "3.safetensors",
|
| 853 |
+
"layers.33.self_attention.to_k.weight": "3.safetensors",
|
| 854 |
+
"layers.33.self_attention.to_k.scales": "3.safetensors",
|
| 855 |
+
"layers.33.self_attention.to_k.biases": "3.safetensors",
|
| 856 |
+
"layers.33.self_attention.to_v.weight": "3.safetensors",
|
| 857 |
+
"layers.33.self_attention.to_v.scales": "3.safetensors",
|
| 858 |
+
"layers.33.self_attention.to_v.biases": "3.safetensors",
|
| 859 |
+
"layers.33.self_attention.norm_q.weight": "3.safetensors",
|
| 860 |
+
"layers.33.self_attention.norm_k.weight": "3.safetensors",
|
| 861 |
+
"layers.33.self_attention.to_out.0.weight": "3.safetensors",
|
| 862 |
+
"layers.33.self_attention.to_out.0.scales": "3.safetensors",
|
| 863 |
+
"layers.33.self_attention.to_out.0.biases": "3.safetensors",
|
| 864 |
+
"layers.33.adaLN_mlp_ln.weight": "3.safetensors",
|
| 865 |
+
"layers.33.mlp.gate_proj.weight": "3.safetensors",
|
| 866 |
+
"layers.33.mlp.gate_proj.scales": "3.safetensors",
|
| 867 |
+
"layers.33.mlp.gate_proj.biases": "3.safetensors",
|
| 868 |
+
"layers.33.mlp.up_proj.weight": "3.safetensors",
|
| 869 |
+
"layers.33.mlp.up_proj.scales": "3.safetensors",
|
| 870 |
+
"layers.33.mlp.up_proj.biases": "3.safetensors",
|
| 871 |
+
"layers.33.mlp.linear_fc2.weight": "3.safetensors",
|
| 872 |
+
"layers.33.mlp.linear_fc2.scales": "3.safetensors",
|
| 873 |
+
"layers.33.mlp.linear_fc2.biases": "3.safetensors",
|
| 874 |
+
"layers.34.adaLN_sa_ln.weight": "3.safetensors",
|
| 875 |
+
"layers.34.self_attention.to_q.weight": "3.safetensors",
|
| 876 |
+
"layers.34.self_attention.to_q.scales": "3.safetensors",
|
| 877 |
+
"layers.34.self_attention.to_q.biases": "3.safetensors",
|
| 878 |
+
"layers.34.self_attention.to_k.weight": "3.safetensors",
|
| 879 |
+
"layers.34.self_attention.to_k.scales": "3.safetensors",
|
| 880 |
+
"layers.34.self_attention.to_k.biases": "3.safetensors",
|
| 881 |
+
"layers.34.self_attention.to_v.weight": "3.safetensors",
|
| 882 |
+
"layers.34.self_attention.to_v.scales": "3.safetensors",
|
| 883 |
+
"layers.34.self_attention.to_v.biases": "3.safetensors",
|
| 884 |
+
"layers.34.self_attention.norm_q.weight": "3.safetensors",
|
| 885 |
+
"layers.34.self_attention.norm_k.weight": "3.safetensors",
|
| 886 |
+
"layers.34.self_attention.to_out.0.weight": "3.safetensors",
|
| 887 |
+
"layers.34.self_attention.to_out.0.scales": "3.safetensors",
|
| 888 |
+
"layers.34.self_attention.to_out.0.biases": "3.safetensors",
|
| 889 |
+
"layers.34.adaLN_mlp_ln.weight": "3.safetensors",
|
| 890 |
+
"layers.34.mlp.gate_proj.weight": "3.safetensors",
|
| 891 |
+
"layers.34.mlp.gate_proj.scales": "3.safetensors",
|
| 892 |
+
"layers.34.mlp.gate_proj.biases": "3.safetensors",
|
| 893 |
+
"layers.34.mlp.up_proj.weight": "3.safetensors",
|
| 894 |
+
"layers.34.mlp.up_proj.scales": "3.safetensors",
|
| 895 |
+
"layers.34.mlp.up_proj.biases": "3.safetensors",
|
| 896 |
+
"layers.34.mlp.linear_fc2.weight": "3.safetensors",
|
| 897 |
+
"layers.34.mlp.linear_fc2.scales": "3.safetensors",
|
| 898 |
+
"layers.34.mlp.linear_fc2.biases": "3.safetensors",
|
| 899 |
+
"layers.35.adaLN_sa_ln.weight": "3.safetensors",
|
| 900 |
+
"layers.35.self_attention.to_q.weight": "3.safetensors",
|
| 901 |
+
"layers.35.self_attention.to_q.scales": "3.safetensors",
|
| 902 |
+
"layers.35.self_attention.to_q.biases": "3.safetensors",
|
| 903 |
+
"layers.35.self_attention.to_k.weight": "3.safetensors",
|
| 904 |
+
"layers.35.self_attention.to_k.scales": "3.safetensors",
|
| 905 |
+
"layers.35.self_attention.to_k.biases": "3.safetensors",
|
| 906 |
+
"layers.35.self_attention.to_v.weight": "3.safetensors",
|
| 907 |
+
"layers.35.self_attention.to_v.scales": "3.safetensors",
|
| 908 |
+
"layers.35.self_attention.to_v.biases": "3.safetensors",
|
| 909 |
+
"layers.35.self_attention.norm_q.weight": "3.safetensors",
|
| 910 |
+
"layers.35.self_attention.norm_k.weight": "3.safetensors",
|
| 911 |
+
"layers.35.self_attention.to_out.0.weight": "3.safetensors",
|
| 912 |
+
"layers.35.self_attention.to_out.0.scales": "3.safetensors",
|
| 913 |
+
"layers.35.self_attention.to_out.0.biases": "3.safetensors",
|
| 914 |
+
"layers.35.adaLN_mlp_ln.weight": "3.safetensors",
|
| 915 |
+
"layers.35.mlp.gate_proj.weight": "3.safetensors",
|
| 916 |
+
"layers.35.mlp.gate_proj.scales": "3.safetensors",
|
| 917 |
+
"layers.35.mlp.gate_proj.biases": "3.safetensors",
|
| 918 |
+
"layers.35.mlp.up_proj.weight": "3.safetensors",
|
| 919 |
+
"layers.35.mlp.up_proj.scales": "3.safetensors",
|
| 920 |
+
"layers.35.mlp.up_proj.biases": "3.safetensors",
|
| 921 |
+
"layers.35.mlp.linear_fc2.weight": "3.safetensors",
|
| 922 |
+
"layers.35.mlp.linear_fc2.scales": "3.safetensors",
|
| 923 |
+
"layers.35.mlp.linear_fc2.biases": "3.safetensors",
|
| 924 |
+
"final_norm.linear.weight": "3.safetensors",
|
| 925 |
+
"final_norm.linear.scales": "3.safetensors",
|
| 926 |
+
"final_norm.linear.biases": "3.safetensors",
|
| 927 |
+
"final_norm.linear.bias": "3.safetensors",
|
| 928 |
+
"final_linear.weight": "3.safetensors",
|
| 929 |
+
"final_linear.scales": "3.safetensors",
|
| 930 |
+
"final_linear.biases": "3.safetensors",
|
| 931 |
+
"final_linear.bias": "3.safetensors"
|
| 932 |
}
|
| 933 |
}
|
vae/0.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 166156496
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a57ecf44fbf13989b4e4d8c2d97cb117832f3ecb53a4f58a027e70f4e65daf66
|
| 3 |
size 166156496
|
vae/model.safetensors.index.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"metadata": {
|
| 3 |
-
"quantization_level": "
|
| 4 |
"mflux_version": "0.18.5"
|
| 5 |
},
|
| 6 |
"weight_map": {
|
|
|
|
| 1 |
{
|
| 2 |
"metadata": {
|
| 3 |
+
"quantization_level": "8",
|
| 4 |
"mflux_version": "0.18.5"
|
| 5 |
},
|
| 6 |
"weight_map": {
|