Simplify model cards (#1204)

We have a lot of unneeded data in the model card - lets just keep the
necessary stuff and add back more data when we need it

EXO still runs! (pipeline on 2)

Co-authored-by: rltakashige <rl.takashige@gmail.com>

dashboard/package-lock.json

@@ -863,6 +863,7 @@
 			"integrity": "sha512-oH8tXw7EZnie8FdOWYrF7Yn4IKrqTFHhXvl8YxXxbKwTMcD/5NNCryUSEXRk2ZR4ojnub0P8rNrsVGHXWqIDtA==",
 			"dev": true,
 			"license": "MIT",
+			"peer": true,
 			"dependencies": {
 				"@standard-schema/spec": "^1.0.0",
 				"@sveltejs/acorn-typescript": "^1.0.5",
@@ -902,6 +903,7 @@
 			"integrity": "sha512-Y1Cs7hhTc+a5E9Va/xwKlAJoariQyHY+5zBgCZg4PFWNYQ1nMN9sjK1zhw1gK69DuqVP++sht/1GZg1aRwmAXQ==",
 			"dev": true,
 			"license": "MIT",
+			"peer": true,
 			"dependencies": {
 				"@sveltejs/vite-plugin-svelte-inspector": "^4.0.1",
 				"debug": "^4.4.1",
@@ -1518,6 +1520,7 @@
 			"integrity": "sha512-LCCV0HdSZZZb34qifBsyWlUmok6W7ouER+oQIGBScS8EsZsQbrtFTUrDX4hOl+CS6p7cnNC4td+qrSVGSCTUfQ==",
 			"dev": true,
 			"license": "MIT",
+			"peer": true,
 			"dependencies": {
 				"undici-types": "~6.21.0"
 			}
@@ -1527,6 +1530,7 @@
 			"resolved": "https://registry.npmjs.org/acorn/-/acorn-8.15.0.tgz",
 			"integrity": "sha512-NZyJarBfL7nWwIq+FDL6Zp/yHEhePMNnnJ0y3qfieCrmNvYct8uvtiV41UvlSe6apAfk0fY1FbWx+NwfmpvtTg==",
 			"license": "MIT",
+			"peer": true,
 			"bin": {
 				"acorn": "bin/acorn"
 			},
@@ -1939,6 +1943,7 @@
 			"integrity": "sha512-fmTRWbNMmsmWq6xJV8D19U/gw/bwrHfNXxrIN+HfZgnzqTHp9jOmKMhsTUjXOJnZOdZY9Q28y4yebKzqDKlxlQ==",
 			"dev": true,
 			"license": "ISC",
+			"peer": true,
 			"engines": {
 				"node": ">=12"
 			}
@@ -2646,6 +2651,7 @@
 			"integrity": "sha512-5gTmgEY/sqK6gFXLIsQNH19lWb4ebPDLA4SdLP7dsWkIXHWlG66oPuVvXSGFPppYZz8ZDZq0dYYrbHfBCVUb1Q==",
 			"dev": true,
 			"license": "MIT",
+			"peer": true,
 			"engines": {
 				"node": ">=12"
 			},
@@ -2833,6 +2839,7 @@
 			"resolved": "https://registry.npmjs.org/svelte/-/svelte-5.45.3.tgz",
 			"integrity": "sha512-ngKXNhNvwPzF43QqEhDOue7TQTrG09em1sd4HBxVF0Wr2gopAmdEWan+rgbdgK4fhBtSOTJO8bYU4chUG7VXZQ==",
 			"license": "MIT",
+			"peer": true,
 			"dependencies": {
 				"@jridgewell/remapping": "^2.3.4",
 				"@jridgewell/sourcemap-codec": "^1.5.0",
@@ -2977,6 +2984,7 @@
 			"integrity": "sha512-jl1vZzPDinLr9eUt3J/t7V6FgNEw9QjvBPdysz9KfQDD41fQrC2Y4vKQdiaUpFT4bXlb1RHhLpp8wtm6M5TgSw==",
 			"dev": true,
 			"license": "Apache-2.0",
+			"peer": true,
 			"bin": {
 				"tsc": "bin/tsc",
 				"tsserver": "bin/tsserver"
@@ -2998,6 +3006,7 @@
 			"integrity": "sha512-+Oxm7q9hDoLMyJOYfUYBuHQo+dkAloi33apOPP56pzj+vsdJDzr+j1NISE5pyaAuKL4A3UD34qd0lx5+kfKp2g==",
 			"dev": true,
 			"license": "MIT",
+			"peer": true,
 			"dependencies": {
 				"esbuild": "^0.25.0",
 				"fdir": "^6.4.4",

pyproject.toml

@@ -17,8 +17,8 @@ dependencies = [
     "loguru>=0.7.3",
     "exo_pyo3_bindings", # rust bindings
     "anyio==4.11.0",
-    "mlx==0.30.3; sys_platform == 'darwin'",
-    "mlx[cpu]==0.30.3; sys_platform == 'linux'",
+    "mlx==0.30.1; sys_platform == 'darwin'",
+    "mlx[cpu]==0.30.1; sys_platform == 'linux'",
     "mlx-lm @ git+https://github.com/AlexCheema/mlx-lm.git@fix-transformers-5.0.0rc2",
     "tiktoken>=0.12.0", # required for kimi k2 tokenizer
     "hypercorn>=0.18.0",

src/exo/worker/engines/mlx/auto_parallel.py

@@ -566,7 +566,7 @@ class MiniMaxShardingStrategy(TensorParallelShardingStrategy):
                 layer.block_sparse_moe.switch_mlp.up_proj
             )
             layer.block_sparse_moe = ShardedQwenMoE(layer.block_sparse_moe)  # pyright: ignore[reportAttributeAccessIssue, reportArgumentType]
-            layer.block_sparse_moe.sharding_group = self.group  # pyright: ignore[reportAttributeAccessIssue]
+            layer.block_sparse_moe.sharding_group = self.group
 
         return model
 
@@ -661,7 +661,7 @@ class GptOssShardingStrategy(TensorParallelShardingStrategy):
             self.all_to_sharded_linear_in_place(layer.mlp.experts.up_proj)
 
             layer.mlp = ShardedGptOssMoE(layer.mlp)  # type: ignore
-            layer.mlp.sharding_group = self.group  # pyright: ignore[reportAttributeAccessIssue]
+            layer.mlp.sharding_group = self.group
 
         return model
 

src/exo/worker/engines/mlx/utils_mlx.py

@@ -169,10 +169,10 @@ def mlx_distributed_init(
 
                 # TODO: update once upstream fixes
                 logger.info(
-                    f"rank {rank} MLX_JACCL_DEVICES: {coordination_file} with devices: {jaccl_devices_json}"
+                    f"rank {rank} MLX_IBV_DEVICES: {coordination_file} with devices: {jaccl_devices_json}"
                 )
                 logger.info(f"rank {rank} MLX_JACCL_COORDINATOR: {jaccl_coordinator}")
-                os.environ["MLX_JACCL_DEVICES"] = coordination_file
+                os.environ["MLX_IBV_DEVICES"] = coordination_file
                 os.environ["MLX_RANK"] = str(rank)
                 os.environ["MLX_JACCL_COORDINATOR"] = jaccl_coordinator
                 group = mx.distributed.init(backend="jaccl", strict=True)
@@ -312,9 +312,6 @@ def get_eos_token_ids_for_model(model_id: str) -> list[int] | None:
     model_id_lower = model_id.lower()
     if "kimi-k2" in model_id_lower:
         return [163586]
-    elif "glm-4.7-flash" in model_id_lower:
-        # 154820: <|endoftext|>, 154827: <|user|>, 154829: <|observation|>
-        return [154820, 154827, 154829]
     elif "glm" in model_id_lower:
         return [151336, 151329, 151338]
     return None

tmp/kill_bridge_plist.sh

@@ -1,84 +0,0 @@
-#!/usr/bin/env bash
-set -euo pipefail
-
-PREFS="${PREFS:-/Library/Preferences/SystemConfiguration/preferences.plist}"
-
-tmpdir="$(mktemp -d)"
-trap 'rm -rf "$tmpdir"' EXIT
-injson="$tmpdir/in.json"
-outjson="$tmpdir/out.json"
-plutil -convert json -o "$injson" "$PREFS"
-
-perl -Mstrict -Mwarnings -MJSON::PP -e '
-  my ($in, $out) = @ARGV;
-
-  open my $fh, "<", $in or die "open $in: $!";
-  local $/;
-  my $txt = <$fh>;
-  close $fh;
-
-  my $json = JSON::PP->new->utf8->relaxed(1);
-  my $d = $json->decode($txt);
-
-  if (ref($d->{VirtualNetworkInterfaces}) eq "HASH"
-      && ref($d->{VirtualNetworkInterfaces}{Bridge}) eq "HASH") {
-    delete $d->{VirtualNetworkInterfaces}{Bridge}{bridge0};
-  }
-
-  my @bridge_svcs;
-  if (ref($d->{NetworkServices}) eq "HASH") {
-    for my $k (keys %{ $d->{NetworkServices} }) {
-      my $svc = $d->{NetworkServices}{$k};
-      next unless ref($svc) eq "HASH";
-      my $iface = $svc->{Interface};
-      next unless ref($iface) eq "HASH";
-      my $dev = $iface->{DeviceName};
-      if (defined $dev && $dev eq "bridge0") {
-        push @bridge_svcs, $k;
-      }
-    }
-    delete @{ $d->{NetworkServices} }{ @bridge_svcs } if @bridge_svcs;
-  }
-
-  my %is_bridge = map { $_ => 1 } @bridge_svcs;
-
-  if (ref($d->{Sets}) eq "HASH") {
-    for my $setk (keys %{ $d->{Sets} }) {
-      my $set = $d->{Sets}{$setk};
-      next unless ref($set) eq "HASH";
-      my $net = $set->{Network};
-      next unless ref($net) eq "HASH";
-
-      if (ref($net->{Interface}) eq "HASH") {
-        delete $net->{Interface}{bridge0};
-      }
-
-      if (ref($net->{Service}) eq "HASH" && @bridge_svcs) {
-        for my $svc (@bridge_svcs) {
-          delete $net->{Service}{$svc};
-        }
-      }
-
-      my $g = $net->{Global};
-      if (ref($g) eq "HASH"
-          && ref($g->{IPv4}) eq "HASH"
-          && ref($g->{IPv4}{ServiceOrder}) eq "ARRAY"
-          && @bridge_svcs) {
-
-        my @so = @{ $g->{IPv4}{ServiceOrder} };
-        @so = grep { !defined($_) || !$is_bridge{$_} } @so;
-        $g->{IPv4}{ServiceOrder} = \@so;
-      }
-    }
-  }
-
-  open my $oh, ">", $out or die "open $out: $!";
-  print $oh JSON::PP->new->utf8->canonical(1)->pretty(1)->encode($d);
-  close $oh;
-' "$injson" "$outjson"
-
-# Convert JSON -> plist (write back as binary1; change to xml1 if you prefer)
-plutil -convert xml1 -o "$PREFS" "$outjson"
-
-# Ask configd to reload SystemConfiguration state
-killall -HUP configd 2>/dev/null || true