get order of seq lengths to make partitions balanced, this is used in balacing sum of seqlength across dp ranks and microbatches Paramete

Return a dictionary containing name to a shape and dtype.

The answer will be considered correct if: (a) it normalizes to the same string as the ground truth answer OR (b) sympy can simplify t

load all model params to cuda

r"""Make an iterator from the DataProto. This is built upon that TensorDict can be used as a normal Pytorch dataset. See https://pytorch.org/t

Make minibatch iterator for updating the actor Args: data (DataProto): a DataProto containing keys ``input_ids``

r"""Works like os.makedirs() but supports hdfs. Super-mkdir; create a leaf directory and all intermediate ones. Works like mkdir, except tha

Exact match of math if and only if: 1. numerical equal: both can convert to float and are equal 2. symbolic equal: both can convert to sy

Normalize the pp vpp params into a complete named parameters. This is useful when gather parameters from pp ranks and passed to a model witho

处理图片并转换为 base64 字符串。 确保图片体积不超过 38KB，同时尊重 max_pixels 限制。 Returns: (base64_str, img, actual_total_pixels)

Repeat the batch data a specified number of times. Args: repeat_times (int): Number of times to repeat the data.

This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch, num_key_value_heads, seqlen, he

Run the VimRAG agent on a sample query. Yields progress events in real-time. Args: sample: Dict with 'query' key

Run main LLM generation loop.

Set ulysses sequence parallel process group.

Split the elements in a tuple/interval, while handling well-formatted commas in large numbers

Update the policy with an iterator of DataProto Args: data (DataProto): an iterator over the DataProto that returns by

Reorder the data on single controller such that each dp rank gets similar total tokens

Bind the worker method to the WorkerGroup

Binds the methods of each worker to the WorkerDict. Note that we only bind public methods that are decorated by register

broadcast tensor in tp shards across mp_group

broadcast tensor in tp shards across mp_group

broadcast tensor in tp shards across mp_group

broadcast tensor in tp shards across mp_group

broadcast tensor in tp shards across mp_group

broadcast tensor in tp shards across mp_group

broadcast tensor in tp shards across mp_group

broadcast tensor in tp shards across mp_group

Build initial messages for the model.

Build the parameter buffer in each pp rank

Check if the resource pool can be satisfied in this ray cluster.

Compose final generation output.

This function should only be called inside by WorkerGroup

Create loss mask for state tokens.

hdfs capable to check whether a file_path is exists

Expands attention_mask from `[bsz, seq_len]` to `[bsz, 1, tgt_seq_len, src_seq_len]`.

Expands attention_mask from `[bsz, seq_len]` to `[bsz, 1, tgt_seq_len, src_seq_len]`.

Expands attention_mask from `[bsz, seq_len]` to `[bsz, 1, tgt_seq_len, src_seq_len]`.