x-transformers 2.0.3__tar.gz → 2.0.4__tar.gz

Sign up to get free protection for your applications and to get access to all the features.
Files changed (55) hide show
  1. {x_transformers-2.0.3 → x_transformers-2.0.4}/PKG-INFO +1 -1
  2. {x_transformers-2.0.3 → x_transformers-2.0.4}/pyproject.toml +1 -1
  3. {x_transformers-2.0.3 → x_transformers-2.0.4}/x_transformers/x_transformers.py +1 -1
  4. {x_transformers-2.0.3 → x_transformers-2.0.4}/.github/FUNDING.yml +0 -0
  5. {x_transformers-2.0.3 → x_transformers-2.0.4}/.github/workflows/python-publish.yml +0 -0
  6. {x_transformers-2.0.3 → x_transformers-2.0.4}/.github/workflows/python-test.yaml +0 -0
  7. {x_transformers-2.0.3 → x_transformers-2.0.4}/.gitignore +0 -0
  8. {x_transformers-2.0.3 → x_transformers-2.0.4}/LICENSE +0 -0
  9. {x_transformers-2.0.3 → x_transformers-2.0.4}/README.md +0 -0
  10. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/all-attention.png +0 -0
  11. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/attention-on-attention.png +0 -0
  12. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/cosine-sim-attention.png +0 -0
  13. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/deepnorm.png +0 -0
  14. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/dynamic-pos-bias-linear.png +0 -0
  15. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/dynamic-pos-bias-log.png +0 -0
  16. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/dynamic-pos-bias-sinusoidal.png +0 -0
  17. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/dynamic-pos-bias.png +0 -0
  18. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/enhanced-recurrence.png +0 -0
  19. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/fcm.png +0 -0
  20. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/ffglu.png +0 -0
  21. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/flash-attention.png +0 -0
  22. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/gate_values.png +0 -0
  23. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/gating.png +0 -0
  24. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/length-extrapolation-scale.png +0 -0
  25. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/macaron-1.png +0 -0
  26. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/macaron-2.png +0 -0
  27. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/memory-transformer.png +0 -0
  28. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/normformer.png +0 -0
  29. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/pia.png +0 -0
  30. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/qknorm-analysis.png +0 -0
  31. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/resi_dual.png +0 -0
  32. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/residual_attn.png +0 -0
  33. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/rezero.png +0 -0
  34. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/rotary.png +0 -0
  35. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/sandwich-2.png +0 -0
  36. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/sandwich.png +0 -0
  37. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/sandwich_norm.png +0 -0
  38. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/scalenorm.png +0 -0
  39. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/talking-heads.png +0 -0
  40. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/topk-attention.png +0 -0
  41. {x_transformers-2.0.3 → x_transformers-2.0.4}/images/xval.png +0 -0
  42. {x_transformers-2.0.3 → x_transformers-2.0.4}/tests/test_x_transformers.py +0 -0
  43. {x_transformers-2.0.3 → x_transformers-2.0.4}/train_copy.py +0 -0
  44. {x_transformers-2.0.3 → x_transformers-2.0.4}/train_enwik8.py +0 -0
  45. {x_transformers-2.0.3 → x_transformers-2.0.4}/train_parity.py +0 -0
  46. {x_transformers-2.0.3 → x_transformers-2.0.4}/x_transformers/__init__.py +0 -0
  47. {x_transformers-2.0.3 → x_transformers-2.0.4}/x_transformers/attend.py +0 -0
  48. {x_transformers-2.0.3 → x_transformers-2.0.4}/x_transformers/autoregressive_wrapper.py +0 -0
  49. {x_transformers-2.0.3 → x_transformers-2.0.4}/x_transformers/continuous.py +0 -0
  50. {x_transformers-2.0.3 → x_transformers-2.0.4}/x_transformers/dpo.py +0 -0
  51. {x_transformers-2.0.3 → x_transformers-2.0.4}/x_transformers/multi_input.py +0 -0
  52. {x_transformers-2.0.3 → x_transformers-2.0.4}/x_transformers/neo_mlp.py +0 -0
  53. {x_transformers-2.0.3 → x_transformers-2.0.4}/x_transformers/nonautoregressive_wrapper.py +0 -0
  54. {x_transformers-2.0.3 → x_transformers-2.0.4}/x_transformers/xl_autoregressive_wrapper.py +0 -0
  55. {x_transformers-2.0.3 → x_transformers-2.0.4}/x_transformers/xval.py +0 -0
{x_transformers-2.0.3 → x_transformers-2.0.4}/PKG-INFO RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: x-transformers
3
- Version: 2.0.3
3
+ Version: 2.0.4
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
{x_transformers-2.0.3 → x_transformers-2.0.4}/pyproject.toml RENAMED
@@ -1,6 +1,6 @@
1
1
  [project]
2
2
  name = "x-transformers"
3
- version = "2.0.3"
3
+ version = "2.0.4"
4
4
  description = "X-Transformers"
5
5
  authors = [
6
6
  { name = "Phil Wang", email = "lucidrains@gmail.com" }
{x_transformers-2.0.3 → x_transformers-2.0.4}/x_transformers/x_transformers.py RENAMED
@@ -1282,7 +1282,7 @@ class Attention(Module):
1282
1282
  dim_kv_input = dim_latent_kv
1283
1283
 
1284
1284
  if exists(latent_rope_subheads):
1285
- assert not exists(rotate_num_heads)
1285
+ assert not exists(rotate_num_heads), '`rotate_num_heads` cannot be set when multi-latent attention is being used'
1286
1286
  rotate_num_heads = latent_rope_subheads
1287
1287
 
1288
1288
  k_dim = dim_head * (kv_heads - latent_rope_subheads)
{x_transformers-2.0.3 → x_transformers-2.0.4}/LICENSE RENAMED
File without changes
{x_transformers-2.0.3 → x_transformers-2.0.4}/README.md RENAMED
File without changes