rxnn 0.1.5__py3-none-any.whl → 0.1.7__py3-none-any.whl

rxnn/rxt/models.py

@@ -2,13 +2,13 @@ import torch
 from torch import nn
 from typing import TypedDict, Union
 from huggingface_hub import PyTorchModelHubMixin
-from rxnn.transformers.positional import RotaryPositionalEmbedding
-from rxnn.transformers.attention import init_attention
-from rxnn.transformers.layers import ReactiveTransformerLayer
-from rxnn.transformers.models import ReactiveTransformerBase, ReactiveTransformerEncoder, ReactiveTransformerDecoder
-from rxnn.transformers.ff import get_activation_layer
-from rxnn.memory.stm import ShortTermMemory
-from rxnn.utils import get_model_size
+from ..transformers.positional import RotaryPositionalEmbedding
+from ..transformers.attention import init_attention
+from ..transformers.layers import ReactiveTransformerLayer
+from ..transformers.models import ReactiveTransformerBase, ReactiveTransformerEncoder, ReactiveTransformerDecoder
+from ..transformers.ff import get_activation_layer
+from ..memory.stm import ShortTermMemory
+from ..utils import get_model_size
 
 
 class RxTAlphaComponentConfig(TypedDict):

rxnn/training/base.py

@@ -7,7 +7,7 @@ from torch.utils.tensorboard import SummaryWriter
 import torch.distributed as dist
 from torch.nn.parallel import DistributedDataParallel
 from typing import Callable
-from callbacks import TrainerCallback
+from .callbacks import TrainerCallback
 
 
 class BaseTrainer(ABC):
@@ -192,7 +192,7 @@ class BaseTrainer(ABC):
                                                epoch * len(dataloader) + batch_idx)
 
                     for callback in self.callbacks:
-                        should_stop = callback.on_batch_end(self.model, batch_idx, loss.item(), batch)
+                        should_stop = callback.on_batch_end(self.model, batch_idx, loss.item() * self.gradient_accumulation_steps, batch)
                         if should_stop:
                             self.is_running = False
 

rxnn/training/bml.py

@@ -5,8 +5,8 @@ import math
 from huggingface_hub import PyTorchModelHubMixin
 from typing import Union
 import torch.distributed as dist
-from rxnn.transformers.models import ReactiveTransformerEncoder, ReactiveTransformerDecoder
-from rxnn.training.base import BaseTrainer
+from ..transformers.models import ReactiveTransformerEncoder, ReactiveTransformerDecoder
+from ..training.base import BaseTrainer
 
 class MLMHead(nn.Module, PyTorchModelHubMixin, license="apache-2.0"):
     def __init__(self, embed_dim: int, vocab_size: int, *args, **kwargs):

rxnn/training/callbacks.py

@@ -3,10 +3,9 @@ import numpy as np
 import torch
 import torch.nn as nn
 from typing import Union
-from rxnn.utils import human_format
 from torch.nn.parallel import DistributedDataParallel
 from huggingface_hub import PyTorchModelHubMixin
-
+from ..utils import human_format
 
 class TrainerCallback:
     def on_epoch_start(self, model: torch.nn.Module, epoch: int) -> None:

rxnn/transformers/attention.py

@@ -2,7 +2,7 @@ import torch
 import torch.nn as nn
 import torch.nn.functional as F
 import math
-from positional import RotaryPositionalEmbedding, RelativePositionalEmbedding
+from .positional import RotaryPositionalEmbedding, RelativePositionalEmbedding
 
 
 class MultiHeadAttention(nn.Module):

rxnn/transformers/layers.py

@@ -1,8 +1,8 @@
 import torch
 import torch.nn as nn
-from attention import MultiHeadAttention
-from ff import FeedForward, GatedFeedForward
-from moe import MoeFeedForward, GatedMoeFeedForward
+from .attention import MultiHeadAttention
+from .ff import FeedForward, GatedFeedForward
+from .moe import MoeFeedForward, GatedMoeFeedForward
 
 
 class ReactiveTransformerLayer(nn.Module):

rxnn/transformers/models.py

@@ -1,8 +1,8 @@
 import torch
 import torch.nn as nn
-from positional import AbsolutePositionalEmbedding
-from mask import create_causal_mask
-from rxnn.memory.stm import ShortTermMemory
+from .positional import AbsolutePositionalEmbedding
+from .mask import create_causal_mask
+from ..memory.stm import ShortTermMemory
 
 
 class ReactiveTransformerBase(nn.Module):

{rxnn-0.1.5.dist-info → rxnn-0.1.7.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: rxnn
-Version: 0.1.5
+Version: 0.1.7
 Summary: RxNN: Reactive Neural Networks Platform
 License: Apache-2.0
 Keywords: deep-learning,ai,machine-learning
@@ -24,8 +24,8 @@ Project-URL: Homepage, https://rxai.dev/rxnn
 Project-URL: Repository, https://github.com/RxAI-dev/rxnn/python
 Description-Content-Type: text/markdown
 
-<img src="logo_rxai.webp" width="300" />
-<img src="logo_rxnn.webp" width="300" />
+<img src="https://github.com/RxAI-dev/rxnn/tree/main/assets/logo_rxai.webp" width="300" />
+<img src="https://github.com/RxAI-dev/rxnn/tree/main/assets/logo_rxnn.webp" width="300" />
 
 # Reactive AI - RxNN
 ## Reactive Neural Networks Platform

{rxnn-0.1.5.dist-info → rxnn-0.1.7.dist-info}/RECORD

@@ -5,25 +5,25 @@ rxnn/memory/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 rxnn/memory/norm.py,sha256=Ofl8Q5NYEF9GQeO0bhM43tkTW91J0y6TSvTAOYMgloM,6278
 rxnn/memory/stm.py,sha256=EsD8slSP4_9dLuq6aFPDmuFe8PWilxh90so5Z3nm-ig,2057
 rxnn/rxt/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-rxnn/rxt/models.py,sha256=NOXEAT262-UVh6q8LjLmdiqOpyze2BkcWTEHoYKAggk,6939
+rxnn/rxt/models.py,sha256=INTFeNcqzAsjyWhNtbBHL4Tx7tYDsaQHgm72tf6u20M,6918
 rxnn/training/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-rxnn/training/base.py,sha256=DHankUIta0wWuczN1O0pK7MCis9dYUvc5dAFkap9t5Q,11030
-rxnn/training/bml.py,sha256=FlMh4_iH67Hsj6kOioeQB5DSqO4qKAlgYsn5OfVmPyc,14581
-rxnn/training/callbacks.py,sha256=BATxW0NrJ3KEpwSquctlJRICKSI0dFYI1MLrji5FdbI,21094
+rxnn/training/base.py,sha256=aZ_1cpCIt5aHI2LiW1QB20s5ltB7H6oDEe0psuLwVIA,11066
+rxnn/training/bml.py,sha256=pyK6aRLpXlPuLge6CQ9PD64Un57yUgbOpu8lUfTdV9k,14575
+rxnn/training/callbacks.py,sha256=IyVJAJ0ggJmfIWBZnpzV9U08URYCeWIStK_wbx7m3pg,21090
 rxnn/training/dataset.py,sha256=vQ5mDF3bA0HXya474n4D4iL8Mn3AEpJukgzFNVkxjGU,5106
 rxnn/training/scheduler.py,sha256=ow6oALzWjWQmHSpcJEjv6tg4g4CDMvr73TypxfcefMc,712
 rxnn/training/tokenizer.py,sha256=4Y41f07uo2KPA_7bp3FCcwGKbXoS2hsckOoXUsXfQxY,8052
 rxnn/transformers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-rxnn/transformers/attention.py,sha256=fqNziF9gZOLM-RnrkaCOkel8w-EkLEAOiv7WO5G6IMw,14021
+rxnn/transformers/attention.py,sha256=FfEYE0THO73p_1eRupr2mcwfW4UbI_riIxkHfr8X_1c,14022
 rxnn/transformers/ff.py,sha256=jJnuBDsnnX5uYC_WZH8cXAYrMnz0P-iX7MwcPivjRtI,2533
-rxnn/transformers/layers.py,sha256=PBlnQzS_cyaAjEAQEgdsttSXhtwWsx4PKxRAtPMhTbY,5361
+rxnn/transformers/layers.py,sha256=jdM7L0uOMO68aZiu9p6jba1Hx3aLGOChF1Zz-j4vJ5U,5364
 rxnn/transformers/mask.py,sha256=J0cfLVLt3SzS2ra3KcY4khrkhI975Dw4CjpUi3Sn25s,419
-rxnn/transformers/models.py,sha256=WP7QoHf3FHbDkc0M5Y69HCLgk2Uwxykt4ScDAAk8EMQ,6717
+rxnn/transformers/models.py,sha256=sLYMkVOWQ1NcM1evpCTUMucXvklySpeNT0IqpIGKmyc,6716
 rxnn/transformers/moe.py,sha256=JQ5QSX4FS7S-fqB7-s1ZmJbPpOeD_Injn8o4vo7wGQE,4936
 rxnn/transformers/positional.py,sha256=2l38RS0Dini3f6Z3LUHr3XwWzg1UK7fO2C6wazWDAYU,4292
 rxnn/transformers/sampler.py,sha256=wSz_1wNloqtuiix5w2Mcsj5NhaO9QlY0j__TVG7wJnM,3938
 rxnn/utils.py,sha256=d5U8i5ukovgDyqiycc2AoxObTz_eF_bgo2MKvdtJ98s,467
-rxnn-0.1.5.dist-info/LICENSE,sha256=C8coDFIUYuOcke4JLPwTqahQUCyXyGq6WOaigOkx8tY,11275
-rxnn-0.1.5.dist-info/METADATA,sha256=9cW2wFwTrdUqpY7J33ap2y8NPKsx7Tnny0fzidhGTGI,14486
-rxnn-0.1.5.dist-info/WHEEL,sha256=fGIA9gx4Qxk2KDKeNJCbOEwSrmLtjWCwzBz351GyrPQ,88
-rxnn-0.1.5.dist-info/RECORD,,
+rxnn-0.1.7.dist-info/LICENSE,sha256=C8coDFIUYuOcke4JLPwTqahQUCyXyGq6WOaigOkx8tY,11275
+rxnn-0.1.7.dist-info/METADATA,sha256=xvzHjAMYR_YDFp_Mb2hFUQe-qQUbhXJ4NNii0w1py7E,14586
+rxnn-0.1.7.dist-info/WHEEL,sha256=fGIA9gx4Qxk2KDKeNJCbOEwSrmLtjWCwzBz351GyrPQ,88
+rxnn-0.1.7.dist-info/RECORD,,