Update deprecated huggingface-cli and fix broken links

docs/README.md

@@ -1,6 +1,6 @@
 # slime Documentation
 
-We recommend new contributors start from writing documentation, which helps you quickly understand SGLang codebase.
+We recommend new contributors start from writing documentation, which helps you quickly understand slime codebase.
 Most documentation files are located under the `docs/` folder.
 
 ## Docs Workflow

docs/en/examples/deepseek-r1.md

@@ -11,12 +11,12 @@ Regarding parallelism, for sglang we will enable EP64, activate dp attention, an
 
 ## Environment Setup
 
-For instructions on setting up the environment and downloading data, please refer to [Example: Qwen3-4B](./qwen3-4B.md).
+For instructions on setting up the environment and downloading data, please refer to [Example: Qwen3-4B](qwen3-4B.md).
 
 To prepare the DeepSeek R1 checkpoint, first you will need to download DeepSeek-R1 to a directory accessible by all machines (hereinafter referred to as `$BASE_DIR`):
 
 ```bash
-huggingface-cli download deepseek-ai/DeepSeek-R1 --local-dir $BASE_DIR/DeepSeek-R1
+hf download deepseek-ai/DeepSeek-R1 --local-dir $BASE_DIR/DeepSeek-R1
 ```
 
 The Hugging Face checkpoint for DeepSeek-R1 is in a block-quantized fp8 format. To convert it into a torch_dist format that Megatron can load, you first need to convert it to a bf16 Hugging Face checkpoint:
@@ -85,7 +85,7 @@ SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" &>/dev/null && pwd)"
 source "${SCRIPT_DIR}/models/deepseek-v3.sh"
 ```
 
-This reads the model's config from [scripts/models/deepseek-v3.sh](../../../scripts/models/deepseek-v3.sh). These configs are all Megatron parameters. When training with Megatron, it cannot read the model config from the checkpoint, so we need to configure it ourselves. We provide some examples in [scripts/models](../../../scripts/models/).
+This reads the model's config from [scripts/models/deepseek-v3.sh](https://github.com/THUDM/slime/blob/main/scripts/models/deepseek-v3.sh). These configs are all Megatron parameters. When training with Megatron, it cannot read the model config from the checkpoint, so we need to configure it ourselves. We provide some examples in [scripts/models](https://github.com/THUDM/slime/tree/main/scripts/models/).
 
 #### CKPT\_ARGS
 

docs/en/examples/glm4-9B.md

@@ -15,14 +15,14 @@ Download the model and data:
 
 ```bash
 # hf checkpoint
-huggingface-cli download zai-org/GLM-Z1-9B-0414 --local-dir /root/GLM-Z1-9B-0414
+hf download zai-org/GLM-Z1-9B-0414 --local-dir /root/GLM-Z1-9B-0414
 
 # train data
-huggingface-cli download --repo-type dataset zhuzilin/dapo-math-17k \
+hf download --repo-type dataset zhuzilin/dapo-math-17k \
   --local-dir /root/dapo-math-17k
 
 # eval data
-huggingface-cli download --repo-type dataset zhuzilin/aime-2024 \
+hf download --repo-type dataset zhuzilin/aime-2024 \
   --local-dir /root/aime-2024
 ```
 
@@ -49,7 +49,7 @@ bash scripts/run-glm4-9B.sh
 
 ### Parameter Introduction
 
-Here, we will briefly introduce the various components of the [run-glm4-9B.sh](../../../scripts/run-glm4-9B.sh) script:
+Here, we will briefly introduce the various components of the [run-glm4-9B.sh](https://github.com/THUDM/slime/blob/main/scripts/run-glm4-9B.sh) script:
 
 #### MODEL\_ARGS
 
@@ -58,7 +58,7 @@ SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" &>/dev/null && pwd)"
 source "${SCRIPT_DIR}/models/glm4-9B.sh"
 ```
 
-Reads the model's config from [scripts/models/glm4-9B.sh](../../../scripts/models/glm4-9B.sh). These configs are all Megatron parameters. When training with Megatron, it cannot read the model config from the checkpoint, so we need to configure it ourselves. We provide some examples in [scripts/models](../../../scripts/models/).
+Reads the model's config from [scripts/models/glm4-9B.sh](https://github.com/THUDM/slime/blob/main/scripts/models/glm4-9B.sh). These configs are all Megatron parameters. When training with Megatron, it cannot read the model config from the checkpoint, so we need to configure it ourselves. We provide some examples in [scripts/models](https://github.com/THUDM/slime/tree/main/scripts/models/).
 
 ⚠️  Ensure that settings such as `--rotary-base` in the model configuration file match the settings of the model you are currently training. This is because different models, even with the same architecture, might use different values. If needed, you can override these parameters in your script after loading the model weights. For instance:
 

docs/en/examples/glm4.5-355B-A32B.md

@@ -5,12 +5,12 @@ This is an example of doing GLM-4.5 RL training using 64xH100 GPUs.
 
 ## Environment Setup
 
-For instructions on setting up the environment and downloading data, please refer to [Example: Qwen3-4B](./qwen3-4B.md).
+For instructions on setting up the environment and downloading data, please refer to [Example: Qwen3-4B](qwen3-4B.md).
 
 First, you will need to download GLM-4.5 to a directory accessible by all machines (hereinafter referred to as `$BASE_DIR`):
 
 ```bash
-huggingface-cli download zai-org/GLM-4.5 --local-dir $BASE_DIR/GLM-4.5-355B-A32B
+hf download zai-org/GLM-4.5 --local-dir $BASE_DIR/GLM-4.5-355B-A32B
 ```
 
 Next, we need to convert the huggingface checkpoint into the torch_dist format with 2 nodes, each with 8 GPUs:
@@ -66,7 +66,7 @@ SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" &>/dev/null && pwd)"
 source "${SCRIPT_DIR}/models/glm4.5-355B-A32B.sh"
 ```
 
-This reads the model's config from [scripts/models/glm4.5-355B-A32B.sh](../../../scripts/models/glm4.5-355B-A32B.sh). These configs are all Megatron parameters. When training with Megatron, it cannot read the model config from the checkpoint, so we need to configure it ourselves. We provide some examples in [scripts/models](../../../scripts/models/).
+This reads the model's config from [scripts/models/glm4.5-355B-A32B.sh](https://github.com/THUDM/slime/blob/main/scripts/models/glm4.5-355B-A32B.sh). These configs are all Megatron parameters. When training with Megatron, it cannot read the model config from the checkpoint, so we need to configure it ourselves. We provide some examples in [scripts/models](https://github.com/THUDM/slime/tree/main/scripts/models/).
 
 #### PERF\_ARGS
 

docs/en/examples/qwen3-30B-A3B.md

@@ -3,7 +3,7 @@
 
 ## Environment Preparation
 
-The environment setup, model download, data, and checkpoint conversion are the same as for the Qwen3-4B model. You can refer to [Example: Qwen3-4B Model](./qwen3-4B.md), replacing mentions of Qwen3-4B with Qwen3-30B-A3B.
+The environment setup, model download, data, and checkpoint conversion are the same as for the Qwen3-4B model. You can refer to [Example: Qwen3-4B Model](qwen3-4B.md), replacing mentions of Qwen3-4B with Qwen3-30B-A3B.
 
 To convert huggingface checkpoint to torch_dist, please try:
 
@@ -29,7 +29,7 @@ bash scripts/run-qwen3-30B-A3B.sh
 
 ### Parameter Introduction
 
-Here, we will briefly introduce the MoE-related parts in the [run-qwen3-30B-A3B.sh](../../../scripts/run-qwen3-30B-A3B.sh) script.
+Here, we will briefly introduce the MoE-related parts in the [run-qwen3-30B-A3B.sh](https://github.com/THUDM/slime/blob/main/scripts/run-qwen3-30B-A3B.sh) script.
 
 1.  To support running Qwen3-30B-A3B in an 8xH800 environment, we need to enable Megatron's CPU Adam to save GPU memory. The corresponding configuration is:
 
@@ -79,7 +79,7 @@ Here, we will briefly introduce the MoE-related parts in the [run-qwen3-30B-A3B.
 slime also supports BF16 training with FP8 inference. For the Qwen3-30B-A3B model, you just need to download the following model:
 
 ```bash
-huggingface-cli download Qwen/Qwen3-30B-A3B-FP8 --local-dir /root/Qwen3-30B-A3B-FP8
+hf download Qwen/Qwen3-30B-A3B-FP8 --local-dir /root/Qwen3-30B-A3B-FP8
 ```
 
 And replace `--hf-checkpoint` with:

docs/en/examples/qwen3-4B.md

@@ -15,14 +15,14 @@ Download the model and data:
 
 ```bash
 # hf checkpoint
-huggingface-cli download Qwen/Qwen3-4B --local-dir /root/Qwen3-4B
+hf download Qwen/Qwen3-4B --local-dir /root/Qwen3-4B
 
 # train data
-huggingface-cli download --repo-type dataset zhuzilin/dapo-math-17k \
+hf download --repo-type dataset zhuzilin/dapo-math-17k \
   --local-dir /root/dapo-math-17k
 
 # eval data
-huggingface-cli download --repo-type dataset zhuzilin/aime-2024 \
+hf download --repo-type dataset zhuzilin/aime-2024 \
   --local-dir /root/aime-2024
 ```
 
@@ -49,7 +49,7 @@ bash scripts/run-qwen3-4B.sh
 
 ### Parameter Introduction
 
-Here, we will briefly introduce the various components of the [run-qwen3-4B.sh](../../../scripts/run-qwen3-4B.sh) script:
+Here, we will briefly introduce the various components of the [run-qwen3-4B.sh](https://github.com/THUDM/slime/blob/main/scripts/run-qwen3-4B.sh) script:
 
 #### MODEL\_ARGS
 
@@ -58,7 +58,7 @@ SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" &>/dev/null && pwd)"
 source "${SCRIPT_DIR}/models/qwen3-4B.sh"
 ```
 
-This reads the model's configuration from [scripts/models/qwen3-4B.sh](../../../scripts/models/qwen3-4B.sh). These are all Megatron parameters. When training with Megatron, it cannot read the model config from the checkpoint, so we need to configure it ourselves. We provide some examples in [scripts/models](../../../scripts/models/).
+This reads the model's configuration from [scripts/models/qwen3-4B.sh](https://github.com/THUDM/slime/blob/main/scripts/models/qwen3-4B.sh). These are all Megatron parameters. When training with Megatron, it cannot read the model config from the checkpoint, so we need to configure it ourselves. We provide some examples in [scripts/models](https://github.com/THUDM/slime/tree/main/scripts/models/).
 
 ⚠️  Ensure that settings such as `--rotary-base` in the model configuration file match the settings of the model you are currently training. This is because different models, even with the same architecture, might use different values. If needed, you can override these parameters in your script after loading the model weights. For instance:
 

docs/en/examples/qwen3-4b-base-openhermes.md

@@ -3,7 +3,7 @@
 
 ## Environment Preparation
 
-First, we need to create a mirror environment and convert the `Qwen3-4B-Base` model by following the [Example: Qwen3-4B Model](./models/qwen3-4B.md).
+First, we need to create a mirror environment and convert the `Qwen3-4B-Base` model by following the [Example: Qwen3-4B Model](qwen3-4B.md).
 
 After that, we will process the SFT data. Here, we use the classic [OpenHermes-2.5](https://huggingface.co/datasets/teknium/OpenHermes-2.5) as an example. First, we process the data into a format suitable for `slime` to load. You can use the following script to add a column that conforms to the OpenAI message format and save it to `/root/openhermes2_5.parquet`.
 
@@ -50,7 +50,7 @@ bash script/run-qwen3-4B-base-sft.sh
 
 ### Parameter Introduction
 
-You can compare [run-qwen3-4B-base-sft.sh](../../scripts/run-qwen3-4B.sh) with [run-qwen3-4B.sh](../../scripts/run-qwen3-4B.sh). You will find that besides changing the model from the instruct version to the base model, the main adjustments are as follows:
+You can compare [run-qwen3-4B-base-sft.sh](https://github.com/THUDM/slime/blob/main/scripts/run-qwen3-4B-base-sft.sh) with [run-qwen3-4B.sh](https://github.com/THUDM/slime/blob/main/scripts/run-qwen3-4B.sh). You will find that besides changing the model from the instruct version to the base model, the main adjustments are as follows:
 
 1.  Removed `SGLANG_ARGS` and `GRPO_ARGS`. This is because it is not necessary to start SGLang or configure GRPO-related settings during the SFT process.
 

docs/en/get_started/qa.md

@@ -49,7 +49,7 @@
 
 9.  **My gradient norm is very high and the training crashes. What should I do?**
 
-    First, ensure that your data and model are compatible. For example, if your data already uses a chat template, check if this template matches the one used by the original model. If the data is correct, please refer to our [Debug Guide](./debug.md) for a more in-depth analysis.
+    First, ensure that your data and model are compatible. For example, if your data already uses a chat template, check if this template matches the one used by the original model. If the data is correct, please refer to our [Debug Guide](../developer_guide/debug.md) for a more in-depth analysis.
 
 10. **My sglang generation takes an extremely long time, GPU power is maxed out, and there's no output for a long while. Why?**
 

docs/en/get_started/quick_start.md

@@ -571,5 +571,5 @@ ray job submit --address="http://127.0.0.1:8265" \
 
 slime has been deeply optimized for distributed training of large-scale Mixture of Experts (MoE) models. We provide some end-to-end training cases for reference:
 
-- [Example: 64xH100 Training GLM-4.5](models/glm4.5-355B-A32B.md)
-- [Example: 128xH100 Training DeepSeek-R1](models/deepseek-r1.md)
+- [Example: 64xH100 Training GLM-4.5](../examples/glm4.5-355B-A32B.md)
+- [Example: 128xH100 Training DeepSeek-R1](../examples/deepseek-r1.md)

docs/en/get_started/usage.md

@@ -67,7 +67,7 @@ MODEL_ARGS=(
 )
 ```
 
-We provide configurations for common models in [scripts/models](../../scripts/models), which you can reuse directly. If you are also using Megatron for pre-training/SFT, you can directly reuse the model configurations from your pre-training/SFT setup.
+We provide configurations for common models in [scripts/models](../../../scripts/models), which you can reuse directly. If you are also using Megatron for pre-training/SFT, you can directly reuse the model configurations from your pre-training/SFT setup.
 
 Note:
 
@@ -99,7 +99,7 @@ Megatron supports several of its custom checkpoint formats. Here are two of the
 
 The `torch` format is Megatron's older storage format. Its structure consists of directories like `mp_rank_xxx`, where each directory corresponds to the checkpoint stored by each rank under a specific parallel partitioning. Because of this, when loading a `torch` format checkpoint, you must ensure that the checkpoint's parallelism strategy matches that of the training task.
 
-We recommend using the `torch_dist` format because it supports automatic parallel sharding, meaning that training tasks with different parallelism settings can share the same checkpoint, which is much more convenient. `torch_dist` is also the default format in the open-source Megatron. A `torch_dist` format checkpoint typically contains a set of `.distcp` files. When using `torch_dist`, you can convert from Hugging Face to `torch_dist` and vice versa using the checkpoint conversion method described in the [README](../../README.md).
+We recommend using the `torch_dist` format because it supports automatic parallel sharding, meaning that training tasks with different parallelism settings can share the same checkpoint, which is much more convenient. `torch_dist` is also the default format in the open-source Megatron. A `torch_dist` format checkpoint typically contains a set of `.distcp` files. When using `torch_dist`, you can convert from Hugging Face to `torch_dist` and vice versa using the checkpoint conversion method described in the [README](../../../README.md).
 
 In terms of storage structure, a Megatron checkpoint typically looks like this, assuming the storage path is `/ckpt/`:
 
@@ -183,7 +183,7 @@ Additionally, we provide a `metadata_key`, which defaults to `"metadata"`. When
 
 slime supports customizing data generation (rollout) to various degrees.
 
-  - By default, it uses the `generate_rollout` function from [slime/rollout/sglang\_example.py](../../slime/rollout/sglang_rollout.py) for data generation. This file implements an asynchronous (asyncio) data generation flow based on SGLang and supports features like dynamic sampling and partial rollout.
+  - By default, it uses the `generate_rollout` function from [slime/rollout/sglang_rollout.py](https://github.com/THUDM/slime/blob/main/slime/rollout/sglang_rollout.py) for data generation. This file implements an asynchronous (asyncio) data generation flow based on SGLang and supports features like dynamic sampling and partial rollout.
 
   - You can completely replace the `generate_rollout` in sglang\_example.py by using the `--rollout-function-path` parameter. You just need to ensure that the function signature passed via `--rollout-function-path` is as follows:
 
@@ -213,7 +213,7 @@ slime supports customizing data generation (rollout) to various degrees.
 
       - `evaluation`: A boolean indicating if the rollout is for evaluation. You can configure a separate evaluation function using `--eval-function-path`.
 
-      - The returned `Sample` type is defined in [slime/utils/types.py](../../slime/utils/types.py). When implementing, you need to ensure the following fields are correctly set:
+      - The returned `Sample` type is defined in [slime/utils/types.py](https://github.com/THUDM/slime/blob/main/slime/utils/types.py). When implementing, you need to ensure the following fields are correctly set:
 
           - `tokens`: The tokens for the prompt + response.
           - `response_length`: The total length of the response. For multi-turn tasks, this is the length of the tokens remaining after the first-turn prompt.
@@ -254,7 +254,7 @@ slime supports customizing data generation (rollout) to various degrees.
         return sample
     ```
 
-    For a more complete version, please refer to [slime/rollout/sglang\_example.py](../../slime/rollout/sglang_rollout.py).
+    For a more complete version, please refer to [slime/rollout/sglang_rollout.py](https://github.com/THUDM/slime/blob/main/slime/rollout/sglang_rollout.py).
 
   - Sometimes, you may also need to support a custom reward model. This can be configured by setting `--custom-rm-path`.
 
@@ -275,7 +275,7 @@ Some parameters related to slime's resource scheduling are configured by slime i
   - `--tp-size` in slime is set using `--rollout-num-gpus-per-engine`.
   - `--model-path` in slime is set using `--hf-checkpoint`.
 
-The way SGLang parameters are integrated into slime can be found in [slime/backends/sglang\_utils/arguments.py](../../slime/backends/sglang_utils/arguments.py).
+The way SGLang parameters are integrated into slime can be found in [slime/backends/sglang_utils/arguments.py](https://github.com/THUDM/slime/blob/main/slime/backends/sglang_utils/arguments.py).
 
 ### How to Use the Router
 
@@ -291,7 +291,7 @@ slime supports different and lightly modified versions of Megatron by reusing co
 
 ### Parameter Configuration
 
-slime directly imports all parameters of the Megatron in the current environment by using `from megatron.training.arguments import parse_args`. If the version of Megatron you are using has parameters defined outside of `parse_args`, you can configure them by passing them in, similar to how it's done in [train.py](../../train.py), for example:
+slime directly imports all parameters of the Megatron in the current environment by using `from megatron.training.arguments import parse_args`. If the version of Megatron you are using has parameters defined outside of `parse_args`, you can configure them by passing them in, similar to how it's done in [train.py](https://github.com/THUDM/slime/blob/main/train.py), for example:
 
 ```python
 if __name__ == "__main__":

docs/en/platform_support/amd_tutorial.md

@@ -85,7 +85,7 @@ Note: We implemented a dedicated AMD conversion script that forces a CPU-only co
 ### Example: Qwen3-4B
 
 We provide examples to use [Qwen3-4B](https://huggingface.co/Qwen/Qwen3-4B), please refer to:
-- [Example: Qwen3-4B Model](../../../scripts/run-qwen3-4B-amd.sh): Just run `scripts/run-qwen3-4B-amd.sh`
+- [Example: Qwen3-4B Model](https://github.com/THUDM/slime/blob/main/scripts/run-qwen3-4B-amd.sh): Just run `scripts/run-qwen3-4B-amd.sh`
 
 ⚠️ TODO: ROCM seems to not support `apex` yet. Thus, we need to disable gradient accumulation fusionby adding the `--no-gradient-accumulation-fusion` flag in the training script currently. We will continue investigating how to enable this.