x-transformers 2.11.16__tar.gz → 2.11.17__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Potentially problematic release.


This version of x-transformers might be problematic. Click here for more details.

Files changed (68) hide show
  1. {x_transformers-2.11.16 → x_transformers-2.11.17}/PKG-INFO +1 -1
  2. {x_transformers-2.11.16 → x_transformers-2.11.17}/pyproject.toml +1 -1
  3. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/free_transformer.py +5 -0
  4. {x_transformers-2.11.16 → x_transformers-2.11.17}/.github/FUNDING.yml +0 -0
  5. {x_transformers-2.11.16 → x_transformers-2.11.17}/.github/workflows/python-publish.yml +0 -0
  6. {x_transformers-2.11.16 → x_transformers-2.11.17}/.github/workflows/python-test.yaml +0 -0
  7. {x_transformers-2.11.16 → x_transformers-2.11.17}/.gitignore +0 -0
  8. {x_transformers-2.11.16 → x_transformers-2.11.17}/LICENSE +0 -0
  9. {x_transformers-2.11.16 → x_transformers-2.11.17}/README.md +0 -0
  10. {x_transformers-2.11.16 → x_transformers-2.11.17}/data/README.md +0 -0
  11. {x_transformers-2.11.16 → x_transformers-2.11.17}/data/enwik8.gz +0 -0
  12. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/all-attention.png +0 -0
  13. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/attention-on-attention.png +0 -0
  14. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/cosine-sim-attention.png +0 -0
  15. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/deepnorm.png +0 -0
  16. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/dynamic-pos-bias-linear.png +0 -0
  17. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/dynamic-pos-bias-log.png +0 -0
  18. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/dynamic-pos-bias-sinusoidal.png +0 -0
  19. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/dynamic-pos-bias.png +0 -0
  20. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/enhanced-recurrence.png +0 -0
  21. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/fcm.png +0 -0
  22. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/ffglu.png +0 -0
  23. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/flash-attention.png +0 -0
  24. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/gate_values.png +0 -0
  25. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/gating.png +0 -0
  26. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/length-extrapolation-scale.png +0 -0
  27. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/macaron-1.png +0 -0
  28. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/macaron-2.png +0 -0
  29. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/memory-transformer.png +0 -0
  30. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/normformer.png +0 -0
  31. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/pia.png +0 -0
  32. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/qknorm-analysis.png +0 -0
  33. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/resi_dual.png +0 -0
  34. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/residual_attn.png +0 -0
  35. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/rezero.png +0 -0
  36. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/rotary.png +0 -0
  37. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/sandwich-2.png +0 -0
  38. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/sandwich.png +0 -0
  39. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/sandwich_norm.png +0 -0
  40. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/scalenorm.png +0 -0
  41. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/talking-heads.png +0 -0
  42. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/topk-attention.png +0 -0
  43. {x_transformers-2.11.16 → x_transformers-2.11.17}/images/xval.png +0 -0
  44. {x_transformers-2.11.16 → x_transformers-2.11.17}/tests/test_x_transformers.py +0 -0
  45. {x_transformers-2.11.16 → x_transformers-2.11.17}/train_belief_state.py +0 -0
  46. {x_transformers-2.11.16 → x_transformers-2.11.17}/train_copy.py +0 -0
  47. {x_transformers-2.11.16 → x_transformers-2.11.17}/train_entropy_tokenizer.py +0 -0
  48. {x_transformers-2.11.16 → x_transformers-2.11.17}/train_enwik8.py +0 -0
  49. {x_transformers-2.11.16 → x_transformers-2.11.17}/train_free.py +0 -0
  50. {x_transformers-2.11.16 → x_transformers-2.11.17}/train_gpt_vae.py +0 -0
  51. {x_transformers-2.11.16 → x_transformers-2.11.17}/train_length_extrapolate.py +0 -0
  52. {x_transformers-2.11.16 → x_transformers-2.11.17}/train_parity.py +0 -0
  53. {x_transformers-2.11.16 → x_transformers-2.11.17}/train_with_muon.py +0 -0
  54. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/__init__.py +0 -0
  55. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/attend.py +0 -0
  56. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/autoregressive_wrapper.py +0 -0
  57. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/belief_state_wrapper.py +0 -0
  58. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/continuous.py +0 -0
  59. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/dpo.py +0 -0
  60. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/entropy_based_tokenizer.py +0 -0
  61. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/gpt_vae.py +0 -0
  62. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/multi_input.py +0 -0
  63. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/neo_mlp.py +0 -0
  64. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/nonautoregressive_wrapper.py +0 -0
  65. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/up_wrapper.py +0 -0
  66. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/x_transformers.py +0 -0
  67. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/xl_autoregressive_wrapper.py +0 -0
  68. {x_transformers-2.11.16 → x_transformers-2.11.17}/x_transformers/xval.py +0 -0
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: x-transformers
3
- Version: 2.11.16
3
+ Version: 2.11.17
4
4
  Summary: X-Transformers
5
5
  Project-URL: Homepage, https://pypi.org/project/x-transformers/
6
6
  Project-URL: Repository, https://github.com/lucidrains/x-transformers
@@ -1,6 +1,6 @@
1
1
  [project]
2
2
  name = "x-transformers"
3
- version = "2.11.16"
3
+ version = "2.11.17"
4
4
  description = "X-Transformers"
5
5
  authors = [
6
6
  { name = "Phil Wang", email = "lucidrains@gmail.com" }
@@ -149,6 +149,7 @@ class FreeTransformer(Module):
149
149
  enc_kwargs: dict = dict(),
150
150
  dec_kwargs: dict = dict(),
151
151
  kl_loss_weight = 1.,
152
+ latent_dropout_prob = 0.,
152
153
  pad_id = -1,
153
154
  **kwargs
154
155
  ):
@@ -187,6 +188,8 @@ class FreeTransformer(Module):
187
188
 
188
189
  self.from_latent_to_condition = nn.Linear(self.binary_mapper.num_codes, dim, bias = False)
189
190
 
191
+ self.latent_dropout = nn.Dropout(latent_dropout_prob)
192
+
190
193
  self.decoder_head = Decoder(
191
194
  dim = dim,
192
195
  depth = dec_head_depth,
@@ -380,6 +383,8 @@ class FreeTransformer(Module):
380
383
 
381
384
  latents, kl_loss = self.encode_to_latents(tokens_for_latents, mask = encoder_mask, per_token_latents = per_token_latents, return_kl_loss = True)
382
385
 
386
+ latents = self.latent_dropout(latents)
387
+
383
388
  condition = self.from_latent_to_condition(latents)
384
389
 
385
390
  # decoder tail